Determinants of concurrent sexual partnerships within stable relationships: a qualitative study in Tanzania.

OBJECTIVE: Concurrent sexual partnerships (CP) have been identified as a potential driver in the HIV epidemic in southern Africa, making it essential to understand motivating factors for engagement in CP. We aimed to assess community attitudes and beliefs about relationship factors that influence men and women in stable relationships to engage in CP in Tanzania. Social exchange theory was used for interpreting the data.
DESIGN: Qualitative study with focus group discussions (FGDs).
SETTING: Semiurban/rural communities in four regions across Tanzania (Dar es Salaam, Shinyanga, Iringa and Mbeya). PARTICIPANTS: 120 women aged 17-45 years and 111 men aged 18-49 years from four study areas participated in 32 FGDs. OUTCOME MEASURES: FGD participants were asked the following questions about CP: definitions and types, motivations and justifications for engaging or not engaging, cultural factors, gender and socialisation, and local resources and efforts available for addressing CP. Our analysis focused specifically on beliefs about how relationship factors influence engagement in CP.
RESULTS: Dissatisfaction with a stable relationship was believed to be a contributing factor for engagement in CP for both men and women. Participants more commonly reported financial dissatisfaction as a contributing factor for women engaging in CP within stable relationships, whereas emotional and sexual dissatisfaction were reported as contributing factors for men and women. Furthermore, participants described how potential outside partners are often evaluated based on what they are able to offer compared with stable partners.
CONCLUSIONS: Efforts to reach men and women in stable relationships with HIV prevention messages must consider the various dimensions of motivation for engaging in CP, including relationship dynamics.

Strengths of this study include investigating an understudied topic in sub-Saharan Africa (relationship dissatisfaction), including participants from multiple regions in Tanzania, and using a theoretical framework to interpret the data.

Limitations of this study include the fact that data come from focus group discussions only, and thus highlight community attitudes and beliefs about concurrent sexual partnerships more than individual experiences.

Introduction

This paper uses the basic tenets of social exchange theory to explore married men's and women's motivations for engaging in concurrent sexual partnerships (CP) in the era of HIV/AIDS. Multiple sexual partnerships have long been recognised as an important factor contributing to the global HIV epidemic, especially in regions such as sub-Saharan Africa where sexual intercourse is the main mode of HIV transmission.1 Many individuals infected with HIV in sub-Saharan Africa are men and women in stable partnerships.1 Some individuals are HIV-positive prior to entering into a stable partnership, which, in turn, places their partner at risk of becoming infected with HIV. Other individuals become infected with HIV through engagement in CP within their stable relationship or through their partner's engagement in CP. In some settings, marriage has been found to be associated with HIV infection among adolescent and adult women and men.2–4 In a study conducted in two high HIV prevalence cities in Kenya and Zambia, it was estimated through deterministic modelling that at least 25% of HIV infections among married men were from concurrent partners outside of marriage.2

CP has gained increasing interest among some HIV experts who believe that it is a key driver in the HIV epidemics of southern Africa.5 6 However, the recognition of the role of CP in the spread of HIV is not new; in the 1990s, Watts and May and Morris and Kretzschmar demonstrated through mathematical modelling that the growth of the HIV epidemic in a population will be much more rapid with high rates of CP compared with serial monogamy.7 8 Recent debate has emerged regarding the validity of the evidence supporting CP as an explanation for HIV epidemics in sub-Saharan Africa. Authors such as Lurie and Rosenthal9 and Sawers and Stillwaggon10 have suggested that there are significant gaps in the evidence in support of the CP argument. It is clear, however, that multiple sexual partners, whether serial or concurrent, increase the risk of HIV infection, and CP does in theory have the potential to lead to even greater rates of HIV transmission as compared with multiple serial partnerships, based on mathematical modelling.7 8 In addition, understanding the correlates of multiple as well as CP is important not only for HIV prevention efforts, but also for other health outcomes such as the prevention of other sexually transmitted infections (STIs) and unintended pregnancies.

The increased attention that CP has received from researchers in Africa in recent years has resulted in a greater understanding of factors related to individual involvement in CP. Several studies throughout Africa have found that community beliefs about the nature of men's strong sexual desires and needs have influenced men's engagement in CP and the social acceptability of this behaviour.11–14 Men's sexual desires, including a desire for sexual variety and adventure, which is frequently believed to be achieved only with more than one woman, has been reported as key reason for engaging in CP.14–16 Furthermore, men have commonly reported engaging in CP because of the sexual unavailability of their wives, especially during times of postpartum abstinence17–20 and migration or time away from home.13 21 22 One study conducted in Uganda found that although marital satisfaction did not deter some men from engaging in CP, marital problems were viewed as a justifiable reason for engaging in CP.12 In this same study, men described CP as a break from household issues such as finances and their partner's complaining.12

For women, engagement in CP has commonly been reported as stemming from a need or desire for financial assistance.15 23–28 Other reasons for engaging in CP described by women in rural Malawi include passion and revenge on an unfaithful partner.24 Young individuals, men and women, in several studies in South Africa explained that CP provides a safety net in case the relationship with a stable partner does not work out.28 29

Many of the studies that have examined factors associated with CP have focused their attention on youth or high-risk groups. Beyond postpartum abstinence, a limited number of studies have looked specifically at individual and couple-level factors that motivate members of stable relationships to engage in CP,12–14 22 30 especially among women.24 Thus, there remains a gap in our understanding of CP in the context of stable relationships. Our study aims to further our understanding in this area by focusing on relationship dissatisfaction within stable relationships and how different dimensions of dissatisfaction (financial, sexual and emotional) are believed to influence one or both partners to engage in CP.

Tanzanian context

This study took place in Tanzania, which is located in East Africa and is home to 43 million people. There are 26 regions in Tanzania and the urban population now makes up one-quarter of the population. Around 1% of men and women aged 45–49 years have never been married, suggesting that marriage in Tanzania is nearly universal.31 Data were collected from four different regions in Tanzania: Dar es Salaam, Shinyanga, Mbeya and Iringa. Dar es Salaam is the largest urban area in Tanzania while the other three regions are mainly rural. Around 35% of the population is Muslim, 30% Christian and 35% indigenous beliefs.32 An estimated 21% of currently married women in Tanzania are living in polygynous unions but this percentage varies greatly by region. In the Tanzania Demographic and Health Survey 2010, no married women in Dar es Salaam and 4–8% of married women in Mbeya, Iringa and Shinyanga reported being in a polygynous union.31

In the final report of the 2011–2012 Tanzania HIV/AIDS and Malaria Indicator Survey (THMIS), some data are presented by geographical zones rather than individual regions. The four regions included in our study were classified under the following zones: Eastern (Dar es Salaam), Lake (Shinyanga), Southwest Highlands (Mbeya) and Southern Highlands (Iringa). Over 90% of men and women in the Eastern zone correctly answered ‘yes’ to the following question: Can people reduce their chance of getting the AIDS virus by having just one uninfected sex partner who has sexual intercourse with no other partners? Only 73–85% of women in Lake, Southwest Highlands and Southern Highlands correctly answered the question, whereas 80–94% of men in these zones did so. The percentage of respondents who had any concurrent sexual partners during the 12 months preceding the survey varied by gender and by region. Fewer than 5% of women in the four zones that included Dar es Salaam, Shinyanga, Mbeya and Iringa had any concurrent sexual partners during the 12 months preceding the survey, whereas men in these zones had much higher rates of concurrent partners ranging from 15% to 19%.33

In the 2011–2012 THMIS, the national HIV prevalence among adults aged 15–49 years was 5%; however, HIV prevalence varied greatly by region. Of the four study regions, Dar es Salaam had the lowest HIV prevalence rate for men (6.9%) and Shinyanga had the lowest rate for women (8.4%). Iringa had the highest HIV prevalence rate for men (9.1%) and Mbeya had the highest rate for women (11%). In all four of the study regions, women had higher HIV prevalence than men.33

Social exchange theory

We utilised social exchange theory to help inform the interpretation of our data on CP within stable relationships from 32 focus group discussions (FGDs) in Tanzania. Social exchange theory emerged in the late 1950s from prominent researchers such as Homans34 and Thibaut and Kelley,35 who utilised ideas from the fields of economics, psychology and sociology.36 Explained by McDonald37 in 1981, social exchange theory is based on the premise that “individuals in social interaction attempt to maximise rewards and minimise costs to obtain the most profitable outcomes.” Social exchange theory has been used in research to understand behavioural outcomes including marital formation and dissolution.36 38

Some social exchange theorists recognise that the success of a relationship is determined by more than just the costs and rewards accrued. Shared expectations regarding the nature of the relationship as well as the expected roles and obligations prescribed for each gender provide a common understanding of what a relationship should resemble within a given sociocultural context.37 These common, socially accepted behaviours and obligations influence the willingness of partners to enter into and continue in a relationship that is rooted in asymmetrical exchanges.37 This point is particularly salient when trying to understand intimate relationships in patriarchal societies where men and women enter into relationships with expectations of distinct gender roles and inequality.

Another essential dimension of social exchange theory is the comparison for alternatives, which Thibaut and Kelley35 define as “the standard the member uses in deciding whether to remain in or to leave the relationship.” Social exchange theorists posit that in addition to the existence of an attractive alternative, diminishing rewards and a breakdown of barriers to leaving the relationship are important factors in relationship dissolution.36 38 For this study, we are not concerned with how these three factors influence the dissolution of stable relationships; rather, we are concerned with how they might influence CP within stable relationships. In societies where marriage is common and lends itself to higher status, it seems logical that a dissatisfying stable relationship with diminishing rewards may lead to infidelity rather than dissolution when a comparatively better relationship is available.

Methods

Study design and data collection

A total of 32 FGDs were conducted in urban and semiurban/rural communities across the four study regions (Dar es Salaam, Shinyanga, Iringa and Mbeya). We were specifically interested in exploring social norms and shared beliefs around CP in HIV-affected areas in order to inform programmatic efforts to reduce HIV transmission; therefore, we purposefully sampled from regions with medium and high HIV prevalence. The FGD method was selected for its ability to obtain greater depth than other methods by creating a dynamic process in which participants can share, build on and disagree with one another on a sensitive topic such as CP.

Stratified purposive sampling was used when recruiting individuals in order to capture younger and older men and women of reproductive age. Having homogenous FGDs by sex and age helps to create an environment in which participants feel more comfortable expressing their opinions and beliefs. Furthermore, stratified purposive sampling facilitates comparison across desired subgroups.39 We hypothesised that the experiences and opinions of CP differ by sex and age. Separating participants by these characteristics allowed us to compare data across these groups in the analysis. Aside from one region (Shinyanga), there were two FGDs for each sex/age category. Owing to an oversight during recruitment in Shinyanga, only one FGD was conducted with younger men while three FGDs were conducted with older men. The intended age group categories were as follows: women 18–29 and 30–39 years, and men 18–29 and 30–44 years. Some individuals outside of the intended age range were mistakenly recruited for the FGDs. As a result, the actual age group categories were: women 17–34 and 24–45 years, and men 15–35 and 20–49 years.

Prior to embarking on data collection, the study team met with local government officials at the selected study sites to introduce the study and its objectives as well as to learn more about the social architecture of the study sites. The officials either personally served or delegated another person to serve as the contact for the study team. The contact person was given the study criteria (age and sex) and then he/she identified potential respondents based on these criteria prior to the visit by the research team. The researchers then met with the potential respondents to assess their fulfilment of the criteria and to invite their voluntary participation in the study.

Eight FGDs were conducted in each of the four regions for a total of 32 discussions. Between 6 and 10 study participants were present at each discussion. Discussions were conducted privately in local venues and varied in length from 40 to 96 min. Each FGD was conducted by a moderator and a note-taker who were Tanzanian research assistants who did not have established relationships with participants prior to the study. Research assistants and participants were not gender matched. The moderators were selected based on a number of desired qualifications, such as having a background in social sciences, previous experience in conducting qualitative research, experience in HIV-related work and the ability to work in difficult circumstances and deal with sensitive topics. All moderators participated in a 3-day training focusing on qualitative data collection methods, reviewing and practising qualitative data collection techniques with study instruments, and research ethics, including confidentiality and informed consent procedures.

FGDs were moderated using semistructured field guides. The following topics on CP were covered in the field guide: definitions and types, motivations and justifications for engaging or not engaging, cultural factors, gender and socialisation and local resources and efforts available for addressing CP. Using the field guide, moderators first asked participants about the different types of sexual relationships that exist in their community followed by questions that were specific to CP. Moderators asked participants about CP by using phrases such as “when a person has more than one sexual partner at the same time.” Field guides were pretested prior to being used for data collection. With study participants’ permission, each discussion was audio-recorded. Participants were provided refreshments during the FGD and were reimbursed for transportation costs.

Data analysis

All audio-recordings were transcribed and translated from Kiswahili into English by field staff in Tanzania. After reviewing the transcripts, it was determined that data saturation was achieved from the 32 FGDs. Participants described their beliefs and opinions of CP within casual and stable relationships but because the reasons for engaging in CP as well as the potential consequences differ by relationship type (casual vs stable), we decided to limit our analysis to one relationship type. As a result, we focused our analysis on the responses that were relevant to community attitudes and beliefs about CP within stable relationships. In some instances, we were able to distinguish the type of stable relationship that participants were describing, such as marriage. In most instances, however, participant responses did not provide sufficient detail to determine the type of stable relationship to which the participants were referring. Responses regarding CP within polygynous unions were excluded from the analysis given the difference between the acceptability and openness of CP within polygynous relationships compared with CP within other types of stable relationships.

A team of researchers developed a coding scheme based on the objectives of the analysis and the relevant themes that emerged from reviewing the transcripts. Table 1 shows 30 codes that were applied across three thematic areas: (1) expectations of committed relationships, (2) motivations for engaging in CP and (3) motivations for not engaging in CP.

Table 1

Codebook for qualitative data analysis

Code	Definition
Thematic area 1: Expectations of committed relationships
EXP-ROLES-GENERAL	Expected roles and behaviour of individuals in a stable relationship (not gender-specific)
EXP-BEHAV-MEN	Expected roles and behaviour of men in stable relationships
EXP-BEHAV-WOMEN	Expected roles and behaviour of women in stable relationships
EXP-FORMATION	Factors that influence people's expectations of stable relationships
EXP-COMMAPPROVE	Community approval of concurrent partnerships within stable relationships
EXP-COMMDISAPPROVE	Community disapproval of concurrent partnerships within stable relationships
Thematic area 2: Motivations for engaging (ME) in concurrent partnerships
ME-SEXUAL	Sexual dissatisfaction, need for sexual variety, sexual desires
ME-ROLEDEFIANCE	Partner defying obligated relationship/marital roles and/or responsibilities
ME-MATERIAL	Desire for material goods
ME-PROBLEMS	Problems with partner in stable relationship
ME-EMOTIONAL	Lack of emotional support (trust, love, etc) from partner in stable relationship
ME-BACKUP	Backup partner in case stable partner leaves relationship
ME-REVENGE	Seek revenge on an unfaithful partner
ME-ALTERNATIVE	An alternative partner exists who is more desirable than stable partner
ME-GENDER	Gender expectation of men to engage in concurrent partnerships
ME-PARENTS	Parent's conversations (or lack of conversations) with their children that encourage or support concurrent partnerships
ME-ECONOMIC	1. If poor: multiple partners can provide added financial support. 2. If well-off: able to afford to have multiple partners
ME-PEER PRESSURE	Indirect or direct pressure received by peers to engage in concurrent partnerships. Direct pressure to have multiple partners; desire to brag to peers about multiple partners. Indirect pressure to have material items that can be obtained via sexual partners
ME-UNSATISFIED	General dissatisfaction with stable relationship that is not explained further
ME-CHARACTER	Habits or desires that are innate and part of a person's character
ME-OTHER	Reasons not covered in any other code
Thematic area 3: Motivation for not engaging (MNE) in concurrent partnerships
MNE-EMOTIONS	Emotional bond with stable partner
MNE-SATISFIED	General satisfaction with stable relationship that is not explained further
MNE-CONSEQUENCES	Fear of negative consequences from stable partner if infidelity is discovered
MNE-RESPECT	Respect for self and/or stable partner
MNE-RELIGION	Religious reasons; faith
MNE-DISEASE	Fear of becoming infected or infecting partner with HIV or STIs
MNE-SETTLED	Settled down, possibly because of age or a family
MNE-BUSY	Busy with other activities (ie, work) leaving no time for additional partners
MNE-OTHER	Reasons not covered in any other code

Thematic coding was completed on all transcripts using Atlas.ti40 by one researcher, who then engaged several other researchers who were familiar with the data in an iterative process of examining and refining the coding scheme and the interpretation of the findings.

On completion of coding, we compiled the quotes for each code from all 32 transcripts. A memo was written for each code describing the patterns and themes that emerged from the coded quotes as well as highlighting gender, age and regional similarities and differences. The salient patterns and themes were further explored using matrices to compare findings across subgroups and to eventually guide the outline of the manuscript.

Different behavioural theories were considered when interpreting findings. Social exchange theory aligned well with the themes and patterns that emerged from the data, and as a result, some of the key concepts of the theory were used to help frame the results. It is important to keep in mind that when individuals in stable relationships have concurrent sexual partners, they are engaging in more than one social exchange. There is the social exchange that takes place with the partner in their stable relationship and the exchange that takes place with their additional partner(s). For this analysis, we focus on the social exchange in the stable relationship and how diminishing or absent rewards from the stable relationship can result in one or both partners engaging in CP. The Results section is organised into four sections that are associated with several basic tenets of social exchange theory: (1) expectations within stable relationships (including role expectations and expected rewards and costs), (2) comparison for alternatives, (3) diminishing rewards and (4) barriers to engaging in CP.

Ethical considerations

The study was described to potential study participants and each provided informed consent prior to participating. Only a few potential participants declined to participate in the study due to time required, monetary expectation and unease with the study topic.

Results

A total of 111 men and 120 women in 32 FGDs participated in the study. All men and women were of reproductive age ranging from 17 to 49 years (table 2). Most focus groups included study participants with varied demographic characteristics and included married and single participants. FGDs with younger participants tended to have fewer married participants whereas FGDs with older participants had more married participants. Common occupations included businessmen and women, vendors, teachers, tailors and farmers. Several participants also reported being students or unemployed.

Table 2

Participant matrix for focus group discussions by age, gender and study location

Region	Women		Men
	Younger number (age range)	Older number (age range)	Younger number (age range)	Older number (age range)
Dar es Salaam	2 (17–24 years)	2 (25–33 years)	2 (23–30 years)*	2 (20–40 years)
Iringa	2 (18–30 years)	2 (27–43 years)	2 (18–33 years)	2 (30–48 years)
Mbeya	2 (18–34 years)	2 (24–34 years)	2 (18–35 years)	2 (25–39 years)
Shinyanga	2 (19–30 years)	2 (25–45 years)	1 (19–26 years)	3 (29–49 years)

*Age range missing for one of the two focus group discussions.

Expectations within stable relationships

According to social exchange theory, a person enters a stable relationship with role expectations that result in certain rewards and costs for each partner. Some of the described expectations of individuals in stable relationships in the FGDs applied to male and female partners such as understanding and loving each other, raising children together and helping each other. Sexual satisfaction and availability were also described as a responsibility of and reward for both partners.

Other role expectations described by participants were gender-specific. The main role of men was commonly described as being the financial provider for the family. Men were also expected to buy their partner’s gifts and other material goods. On the other hand, gender-specific roles for women were mostly focused on fertility and taking care of the home. For example, participants from an FGD of older females described how men look for an additional partner if their wife does not become pregnant within 6 months of marriage. Participants also described an unequal distribution of power in relationships, in part due to cultural practices such as the payment of a bride price, which has resulted in the expectation that women are to be obedient to the demands of their partner and accept their partner's poor behaviour, including infidelity.

One partner's inability or unwillingness to fulfil his/her expectations could decrease the rewards for the other partner, which, as described by several participants, can result in relationship dissatisfaction. Furthermore, participants’ responses, including the quote below, suggest a common belief that when a person experiences relationship dissatisfaction, they are more likely to engage in CP as a means to obtaining their expected or desired rewards:

If he the man is married to a woman who doesn't care about him, a woman who doesn't cook and who doesn't stay at home most of time [he] would get tempted to be with another woman who would show even a little bit of concern for him. (Younger female, Shinyanga)

On the other hand, relationship satisfaction was described as a protective factor against engaging in CP. Participants explained that individuals who are satisfied with their relationship and who have their needs and desires met are less likely to engage in CP. The following quote, which resembles several other quotes, was given in response to a question about why some people choose not to engage in CP.

What I can say is that when you are satisfied with what you have then you can't do strange things like finding someone outside of the relationship. (Older Female, Dar es Salaam)

Comparison for alternatives

Social exchange theorists posit that the presence of an attractive alternative partner contributes to relationship failure.36 There were many examples from the FGDs in which participants described how they believe that individuals in their community compare their stable partner with alternative partners with regard to the satisfaction that could be achieved with alternative partners or what the alternative partners could provide that the stable partner cannot.

Maybe her man doesn't provide her with everything she needs. So, whenever she gets a man who can fill the gap she will be with that man while still going on with her man. (Older Female, Mbeya)

On the contrary, satisfaction within one's own relationship makes alternative options seem less meaningful and attractive. Several participants, such as the participant in the quote below, described how relationship satisfaction or feelings of being ‘settled’ are believed to dissuade individuals from considering alternative partners.

Maybe my sexual partner cares for me in everything I need, he loves me, you see, so there is no need of having multiple sexual partners, I am satisfied being with this man. (Younger Female, Dar es Salaam)

Diminishing rewards in stable relationships

Another factor that social exchange theorists believe contributes to relationship failure is diminishing rewards.36 As noted in a previous section, decreasing rewards in a relationship, often due to unfulfilled role expectations, can result in relationship dissatisfaction. Participants explained that relationship dissatisfaction is a key factor in individuals deciding to engage in CP in their communities. Participants described many aspects of dissatisfaction; however, dissatisfaction with one's financial/material situation, sexual aspects and/or emotional quality of their relationship were the most commonly discussed factors with regard to engagement in CP within stable relationships. Each of these three types of dissatisfaction is expanded below.

Financial and material dissatisfaction

There was a shared expectation among most participants that a man's primary role within a stable relationship is to be the material provider for his partner and family. There was some belief among study participants that if men do not fulfil their role as provider, then women may seek financial and material assistance for themselves and their children from an alternative partner; however, there was some variation in responses by subgroup. Older men, especially from Shinyanga and Mbeya, less commonly mentioned financial and material dissatisfaction as a contributing factor for engaging in CP in stable relationships compared with the other subgroups.

On the basis of participants’ discussions, the economic opportunities for women appear to be limited in this context. A lack of income increases a woman's dependence on her partner or another man for financial assistance. Female participants were more likely than male participants to state that the financial hardships of women can motivate women in stable relationships to engage in CP. In the quote below, a participant, in response to the moderator's question about reasons women have more than one sexual partner at a time, asserts that financial insecurity is a key factor:

I think it is all in our lives in the Sukuma society [Tribe in Northwest Tanzania]. Men do not really care about inside needs, it is an obligation of the woman to seek for an alternative way to survive with the kids, and without doing so she won't be able to take care of the kids. She has to come up and go elsewhere to look for other means. (Older Female, Shinyanga)

Furthermore, a male participant in another FGD described how a man's mistress receives more money than his wife at home, tempting even married women to become someone else's mistress:

The woman outside is the one who gets more money. When you have a woman outside she gets more money than the wife at home. So even your wife, when she meets someone else's husband he will give her more money than what he leaves at home. Things like this contribute [to women having concurrent partnerships], that is why you find that women are very difficult. (Younger Male, Dar es Salaam)

Although not mentioned as frequently as for women, participants in three of the four study regions explained that some men are tempted to engage in CP within stable partnerships because of financial and/or material incentives offered by another woman. Participants, like the one quoted below, frequently used the word ‘trap’ when describing how they view experiences of being approached by a potential partner who is financially secure. Using the term ‘trap’ to describe experiences suggests a perceived inability to avoid CP.

A man might have a wife but his financial situation is not good, and he might find a wealthy woman who wants to take care of him and his wife. He sits down and thinks, “if I go there I will get suits that I have never worn before and she gives me some money to feed my wife,” as result he is trapped. (Older Male, Iringa)

Sexual dissatisfaction

Sexual desires and dissatisfaction of both genders were among the most commonly believed reasons for why individuals engage in CP within stable relationships, with some variation across subgroups. Women tended to discuss women's sexual desires and dissatisfaction more than men's sexual desires and dissatisfaction and vice versa in all regions except Iringa. Furthermore, younger participants and participants from Dar es Salaam discussed sexual desires and dissatisfaction more frequently than older participants and participants from other regions.

Sexual issues that emerged from the FGDs ranged from sexual infrequency to lack of variety to physical features such as the size of a woman's buttocks or a man's penis. Participants in several FGDs explained that a disagreement about the timing and frequency of sex within a relationship can provoke one or both members to look for sexual satisfaction elsewhere. In these descriptions, it was more commonly women than men who were charged with being sexually unavailable to their partners. Men were often described as being unable to control their sexual desires, especially when their partner was sexually unavailable. Other participants frequently discussed how the inability of an individual to sexually satisfy his/her partner ‘forces’ the partner to seek an outside partner, as if there is no other choice:

For a man you might find he is having a wife at home; it may happen that at the time when a man needs sex his wife is not ready. Therefore, it will force him to look for somebody else who will quench his needs. (Younger Female, Shinyanga)

Reasons for sexual dissatisfaction included not only timing and frequency, but also sexual variety and practices. Several participants, especially male participants, equated monogamous sexual relationships to eating the same meal every day and believed some men could not be satisfied by one woman. Sexual variety was believed to be satisfied by adding sexual partners. Participants explained that if certain desired sexual practices are not performed within a stable relationship, then the sexually dissatisfied partner might try to find someone outside of the partnership to perform the desired acts. The exchange below between a study participant and a moderator highlights how anatomical disadvantages in one partner may provoke the other partner to engage in CP:

Study Participant: In order for them [sexual partner] to hide their physical disabilities they would ask you to tolerate with them. For me to show him that I really can be patient, I would stay while in reality I have someone on the outside that satisfies me. Therefore with this one, even if all we do is just snore together in bed, I have covered up for him because people know we are together though there isn't much going on between us… but the truth is that I get my services from elsewhere.

Moderator: What services are you referring to?

Study Participant: That it is officially known that the man I have sex with is my husband but because of his sexual problems of having a small penis and can't satisfy me then I am forced to go and look at someone else who can satisfy me… but I won't leave him because if I do then he will be embarrassed. (Older Female, Shinyanga)

Emotional dissatisfaction

According to responses provided by study participants, people need and seek more than just financial and sexual fulfilment to be wholly satisfied in a relationship. Several participants described how a lack of emotional fulfilment in a stable relationship could impel a person to engage in CP. Participants described how dissatisfaction with different emotional aspects of a relationship could lead to infidelity with no evident distinction between regions; however, there were some differences between gender and age. Female participants and younger participants stated that emotional issues within stable relationships motivate individuals to engage in CP more often than male participants and older participants.

The absence of love in a relationship, a lack of trust between partners and relationship ‘misunderstandings’ or ‘quarrels’ were described as reasons individuals within stable relationships engage in CP. The participant quoted below demonstrates how a lack of emotional fulfilment in a stable relationship can motivate a person to seek fulfilment from an outside partner:

I have a vivid example. I have my sister, just a youth, who has misunderstandings with her husband but she is having an affair. It is because they are having misunderstandings. She stays just because they are married but to get sex she is having another man. (Younger Male, Dar es Salaam)

The dissolution of love in a stable relationship was described as another facilitating factor in one or both partners’ decision to engage in CP. The example provided by one study participant explains how the birth of a child can shift a woman's love and attention from the husband to the baby:

You may find maybe the woman after giving birth, her love shifts to her children. She reduces [the love she gives] to you, so you find a man looks for another place so that he can at least be loved. (Older Male, Mbeya)

Barriers to engaging in CP

Social exchange theorists posit that in addition to diminishing rewards and the presence of an attractive alternative partner, a breakdown of barriers to exit a relationship also contributes to relationship failure.36 Given that our outcome of interest was the engagement of CP within stable relationships, we explored beliefs about how certain barriers inhibit the risk of engaging in CP within stable relationships and how these barriers differ for men and women.

Commonly mentioned barriers to engaging in CP include the risk of HIV infection, religious objections, potential consequences from one's stable partner and community disapproval. The majority of participants who mentioned the risk of HIV as a barrier to engaging in CP came from Iringa. No indication was given that this particular barrier was weakening. In fact, one participant noted that as people have become more educated about HIV/AIDS and know friends infected with HIV, their fear of HIV has grown and has caused them to realise the importance of staying faithful to their partner. Religious objection, which was mentioned by male and female participants from all four regions, was another commonly mentioned barrier to engaging in CP. Participants noted Islam as an exception given the acceptance of polygamy in the religion.

Another barrier to engaging in CP is the potential consequence if the partner finds out about the infidelity. In the male and female FGDs, participants explained that the consequences for engaging in CP are much greater for women than men. According to many participants, if a woman in a stable relationship is caught being unfaithful, her partner is likely to end the relationship, whereas if a man in a stable relationship is unfaithful, his partner is expected to accept the infidelity and forgive him:

In past years it was normal for men to have multiple sexual partners but not for women. For example, it is so difficult to forgive my wife after seeing her having sex with another man but my wife is likely to forgive me if I do the same thing. Now you can see that we have done the same thing but a woman forgives, not a man. In very rare cases you can see a man forgiving his wife in such circumstances. (Young Male, Mbeya)

The younger and older participants across all four sites noted disapproval of CP within their communities. Many participants used terms such as ‘unacceptable,’ ‘immoral,’ and ‘not normal’ to describe their opinions about CP. It was evident that CP is less acceptable for women than men. Despite the fact that many participants noted community disapproval of CP, others disagreed with their peers and suggested that there is community acceptance of CP, especially for men. Several participants noted that CP is often seen as a privilege or even a man's right.

There were some participants who felt that it was acceptable for women to engage in CP, especially for survival purposes. Furthermore, it was revealed in several FGDs that women in certain tribes are encouraged to engage in concurrent partnerships. For example, the participant below describes an initiation ceremony she witnessed on TV:

The girl was asked what are they taught in those ceremonies and she said that they are told to have mafiga matatu which means at least three male sexual partners. Because if you have three men, one may not have money, but not all of them in the same day. (Young Female, Dar es Salaam)

Despite participants’ differing views on community attitudes about CP, many agreed that CP is common and even expected. The discussions from the FGDs suggested a general tolerance for men and, in some cases, women to engage in CP.

Discussion

Understanding the factors that motivate individuals in stable relationships to engage in CP is important for HIV prevention as well as other outcomes, including the prevention of other STIs and unintended pregnancies. Social exchange theory has been widely used in research on marriage and attributes the lack of rewards in a relationship, a breakdown of barriers to leave a relationship and the existence of an attractive alternative to marital dissolution.38 In agreement with social exchange theory, we found that similar criteria can be applied to the engagement of CP within stable relationships.

We described the expectations of men and women in stable relationships in Tanzania and the perception that when expectations are not met and benefits diminish, relationship dissatisfaction ensues. We found that relationship dissatisfaction is believed to be a key motivating factor for engaging in CP within stable relationships in Tanzania, especially when an alternative partner is present and is perceived to provide something that is lacking in the stable relationship. Furthermore, potential barriers to engaging in CP were discussed, and there was evidence that some barriers are more significant for women than men, such as consequences from one's partner and community disapproval of CP. Engagement in CP is one of many possible outcomes of relationship dissatisfaction and is likely to be a more common outcome in communities that are accepting of CP.

In an application of social exchange theory to marital relationships, the authors note that it is not the accrual of costs that theorists believe contributes to marital dissolution but rather the diminishing rewards and a breakdown of barriers.36 This might explain why participants in our study focused on the rewards, or lack of rewards, in relationships when discussing motivating factors for engaging in CP rather than high or increasing costs endured in a stable relationship.

Financial, sexual and emotional dissatisfaction in stable relationships were three common dimensions of dissatisfaction discussed by participants. Consistent with findings from previous studies in Southern Africa,23–27 we found that financial and material assistance are common factors believed to be related to women's decision to engage in CP within stable relationships. For some women, engagement in CP is believed to be a means of survival. Women whose partners are unable or unwilling to financially provide for them may see no other alternative than to seek assistance from another man. For other women, it is believed to be less about survival and more about material desires. Either way, a lack of economic opportunities available to women in Tanzania appear to be contributing to women engaging in CP with men who have more economic means than their stable partner. Implementing programmatic efforts aimed at improving the economic opportunities for women would offer women alternative options to satisfying their financial needs and desires. Male partners often control the earnings of women in Tanzania, thus programmes that aim to improve women's economic opportunities should also address issues of gender inequality.

Our analysis showed that participants believed sexual dissatisfaction to be a more common reason for men to engage in CP than women. These findings are consistent with other research that has been conducted in the region about beliefs of men's sexuality and sexual desire.11 12 15 21 Although discussed less frequently than men's sexual satisfaction, we found that participants commonly discussed women's sexual satisfaction as a reason for women to engage in CP within stable relationships. This provides insight into an area of sexual health that has been largely ignored: sexual pleasure. This is especially true in the developing world and even more so for women. These findings contribute to our understanding of women engaging in CP within stable relationships and calls for more research on the sexual pleasure and satisfaction of women in sub-Saharan Africa.

In our analysis, men and women were described as seeking outside partners to improve and broaden their sexual experiences. Similar results were found in a study in Zimbabwe by Kesby.16 Sexual dissatisfaction may lead to risky sexual behaviour that puts individuals and their partners at risk for HIV and other STIs, especially if condoms are not being used with outside partners. There is a lack of understanding with regard to how common sexual displeasure is and how partners cope with this displeasure. We hope that researchers will work to further understand sexual relationships and their link with sexual risk behaviours by using quantitative research methods to better quantify different dimensions of sexuality among married couples including sexual satisfaction and sexual frequency, and by using qualitative research methods to examine how couples conceptualise sexual satisfaction and how they cope with sexual dissatisfaction. This research would improve our ability to design appropriate sexual and reproductive health programmes and services for men, women and couples.

Emotional aspects of relationships are also often overlooked in research and practice throughout Africa; yet our findings suggest a need to give more attention to these factors. Participants frequently described ways in which they believe emotional quality of relationships can contribute to men and women engaging in CP. Understanding how the quality of relationships such as love, trust and commitment influences engagement in CP is pertinent for developing programmes and policies that recognise and address the human dimension of sexual decisions and behaviours.

Limitations

There are some limitations to applying the social exchange theory to the engagement of CP within stable relationships. Individuals engaged in CP are involved in concurrent exchanges; however, social exchange theory does not fully consider the interaction of these exchanges. Furthermore, social exchange theory is based on rational actor assumptions, yet we found evidence that engagement in CP within stable relationships may involve some emotion-based decision-making. Another limitation to this theory is the lack of recognition that it gives to structural factors including gender and economic inequalities that may motivate behaviour. Future research on CP should consider combining elements of a gender theory such as the Theory of Gender and Power41 with the Social Exchange Theory.

There are other limitations to our study that should be considered when drawing conclusions about our findings. Data were collected to understand multiple aspects of CP including factors that motivate individuals to engage in CP. Study participants were asked to discuss these motivating factors and often spoke specifically about CP within stable relationships, especially marriage. Participants were not, however, asked explicitly to describe CP within stable relationships. Asking participants directly about CP within stable relationships may have resulted in more detailed descriptions and/or additional motivating factors that were not captured from our study. In addition, doing so would have allowed us to examine and compare motivating factors for engaging and not engaging in CP between specific types of stable relationships, such as unmarried and married. More research is needed to explore this specific area of CP.

Another limiting factor is that our analysis is based on FGDs. The FGDs were intended to provide data on perceived social norms, attitudes and behaviours of the communities sampled, thus participants were not asked to share their own experiences. As a result, some of the quotes given by participants may be exaggerated because participants were speaking in general terms and not about personal experience. Gathering data on individual experiences of CP within stable relationships through in-depth interviews and comparing the results to the community attitudes and beliefs described in our study is an important next step.

Conclusion

This is one of the first analyses, to our knowledge, that has specifically examined the role of relationship dissatisfaction in CP within stable relationships in Tanzania. Our results support the need for further research and programmatic efforts in this area. A better understanding of how couples communicate, interact and resolve problems, especially with regard to sexual health, is needed in order to adequately develop policy and programmatic responses. Helping individuals to obtain and maintain healthy, satisfying relationships might have a profound impact on the health of individuals, their partners and their families.