Fertility Intentions and Use of Contraception among Monogamous Couples in Northern Malawi in the Context of HIV Testing: A Cross-Sectional Analysis

Introduction

Knowledge of HIV status could influence desired fertility in two opposing ways [1]. Knowledge of HIV positive status could reduce the desire for additional children through concern for the health consequences for the mother and child and ability to care for the children until maturity. Alternatively, awareness of HIV status could accelerate childbearing for fear of not reaching a desired family size before the onset of AIDS or due to an increased desire for more children because of high levels of child mortality [2], [3], [4].

In the mid 1990s testing for HIV was not widespread and an HIV diagnosis late in the course of disease implied that many infected people could not survive long enough to change their fertility [3]. As a consequence, studies on the effect of HIV on fertility found that although being HIV positive is associated with lower fertility, knowledge of HIV positive status did not translate into a noticeable negative or positive impact on fertility [1], [3].

Research to date on knowledge of HIV status and couples’ fertility intentions has been limited. Individual level studies reveal that HIV positive individuals want to limit childbearing [5], [6], [7] in consideration of factors such as risk of infection to their partners and their unborn children [8]; awareness that the onset of AIDS will lead to ill-health and increased difficulties in caring for themselves and their children [9]; perceived community disapproval associated with HIV and reproduction even though motherhood and childbearing are socially valuable for young HIV positive people [10]. These factors have varied by gender with women expressing more concern about the health consequences of carrying a pregnancy to full-term and childbearing while men are more concerned about their own early death and the future of their children [11]. Men and women who tested positive and yet had perceived themselves as being HIV negative changed their fertility intentions from desiring to have more children to desiring to stop [12].

While research findings have been consistent on the effect of HIV awareness on fertility intentions, studies of the effect of antiretroviral treatment (ART) on fertility intentions have tended to be inconclusive [8], [13], [14]. Studies in Uganda and South Africa found that women on ART were more likely to want more children compared to women who were not on ART and that this desire increased with duration of being on ART [13], [14] But other studies found no clear or no evidence that being enrolled in an ART treatment clinic modified fertility intentions [5], [9].

Few studies have analysed the relationship between fertility intentions and HIV awareness in couples and how spousal differences in fertility intentions influence use of contraception. A review of these studies found mixed and inconsistent evidence of which partner carries more weight in reproductive decisions [15]. Another review argued that couple’s joint fertility intentions lead to better predictions of subsequent behaviour [16]. Doodo [17] analysing two rounds of demographic and health survey data from Ghana and Kenya compared reproductive intentions of men and women; he found little evidence that men’s preferences are more influential. A comparative study in 18 countries using couples’ data from the Demographic and Health Survey found that there is a process of negotiation between couples with different reproductive goals and that use of modern methods of contraception was higher when both want to stop, except in Malawi where use was higher when the man, but not the wife, wanted to stop. [18].

This study, using data on fertility intentions and information on HIV status of couples linked to an on-going demographic surveillance system in Karonga district in Northern Malawi will investigate how knowledge of HIV status alters men’s, women’s and couples’ fertility intentions and how existing differences in fertility intentions influence use of contraception. We achieve this by answering the following specific research questions:

1. How does knowledge of HIV status influence fertility intentions for men, women and couples?
2. How does knowledge of HIV status influence contraceptive use?

Methods

This study uses data collected between October 2008 and September 2009 from a module on fertility intentions linked to an on-going Demographic Surveillance system (DSS). The DSS baseline census was conducted in 2002–2004 in a population of around 33,000 individuals, following which the population has been under continuous surveillance with an annual re-census. A population-based adult HIV and behaviour survey started in the DSS area in September 2007 [19], [20].

Questions, asked separately to men and women, using the local language (Tumbuka), captured marital status, current fertility, including total number of children ever born and surviving (separating those born with the current spouse and those born from past unions), fertility intentions, including perception of their partner’s desire and the preferred timing of the next birth, and use of contraception. Respondents who wanted no more children were also asked consequences of having another child (as an open question). Unique individual identification numbers in the DSS permits the linking of couples’ intentions data. Interviews were conducted by trained interviewers, one-to-one and in private.

This paper analyses the baseline round of data (2008–2009) that is part of three rounds of data collection: 2008–2009, 2009–2010, and 2010–2011 approved by the National Health Research Ethic Committee of the National Commission for Science and Technology of Malawi. Data on fertility were collected along with data on sexual behaviour as part of a survey which also offered HIV testing. Individuals could refuse testing and still complete the questionnaire. The data were collected approximately one month after the re-census of that area, which included collection of socio-economic data and identification of spouses.

We collected data on HIV status between October 2007 and October 2008 using door-to-door HIV testing with rapid tests one year before the onset of the fertility intentions study [20]. We collected data on knowledge of HIV status and whether the participants have ever been enrolled on ART after filling fertility intentions questions. Free ART has been available in Karonga District since 2005 and in the DSS area since 2006. For the analysis we used stated knowledge of HIV status rather than actual status, since it is knowledge that will influence behaviour.

Ethics

Ethics approval for the study was received from the Health Sciences Research Committee, Malawi, and the ethics committee of the London School of Hygiene & Tropical Medicine, UK. Before the start of the demographic surveillance the Traditional Authority that covers the area, and all village headmen and traditional advisors in the study area were informed about the aims of the study and the nature of the data to be collected, and their approval and verbal consent was sought. All household members were given a similar explanation and interviews were only conducted if verbal consent was given by the household head and by the respective household members. The consent for the demographic surveillance was recorded by the interview sheet being filled. Refusals were recorded in field registers. During the baseline census 15 households did not provide verbal consent and have consequently been excluded from the study. The socio-demographic data for this study come from the basic demographic surveillance for which the ethics committees agreed that written consent was not needed. For the fertility intentions survey and for HIV testing, separate individual written consent was sought.

Statistical Analysis

We used Stata 11 software to analyse data. We carried out bi-variate analysis using Chi-square tests for associations and then ran logistic regression analyses for the probability of wanting no more children. We examined whether fertility intentions of couples varied by their HIV status. There were four possible outcome combinations for the couples: both wanting more children or undecided; both wanting no more children; wife wanting no more children husband wanting more/undecided; husband wanting no more children wife wanting more/undecided. The undecided were grouped with those who wanted more children because the undecided tend to be classified the same as the majority among those in the survey with stated desires [21]. Owing to small numbers, we subsequently collapsed this outcome variable to two categories: both want no more children and both or either want more or undecided. There were also four different ways in which knowledge of HIV status could be distributed in the couple: concordant positive; concordant negative; discordant husband positive, discordant wife positive. Using logistic regression we also explored how couples’ knowledge of their HIV status influences use of contraception. We used women’s reports of use of contraception to represent couples’ use of contraception.

Results

Of 4654 married couples known from the census, there were 2748 couples where both husband and wife were seen agreed to take part in the fertility intentions study. We dropped 112 records due to mismatching information on the partner’s identity, marital status or the name of the wife or husband. We obtained information on 2636 matched unions of which 67% were monogamous couples and not currently pregnant. We excluded 11%who were pregnant, 20% polygamous and 2% who were outside childbearing age group. Figure 1 presents information on how couples were included into the analysis, their Knowledge of HIV status and use of ART.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image ()
• TIFF
  original image ()

Figure 1. Monogamous married couples’ inclusion, HIV and ART status chart.

doi:10.1371/journal.pone.0051861.g001

The mean age of women was 29 years (range 15–49 years) with HIV positive women being slightly older (mean age 33 years range 17 to 48) than HIV negative women (mean age 29 years, range 15 to 49). For the men the mean age was 35 years (range 17–66 years) with HIV positive men being older (mean age 43 years range 25 to 64) than HIV negative men (mean age 35 years range 17 to 66 ). The mean duration of marriage was 9.8 years with a range of 0 to 35 years. 92/1766 men (5%) knew that they were HIV positive and 66/1766 women (4%) knew that they were HIV positive. HIV testing at the time of the survey revealed that an additional 17 women and 7 men were HIV positive. Based on previous test results from sero-surveys and others studies carried out among the study population, 98% of those who reported being HIV negative were really HIV negative at the time they last had an HIV test before the survey (1139/1160 men and 1318/1341 women with results available). 100% (64/64) of those who reported being HIV positive were indeed HIV positive. 35/51 HIV positive women reported to have disclosed their HIV status to their husbands and 59/82 HIV positive men reported to have disclosed their HIV status to their wives. The proportions were similar for HIV negative men and women: 68% (866/1262) and 69% (890/1293) respectively had disclosed.

Table 1 shows the proportion of men and women wanting no more children, and proportion of women using contraception, by background characteristics of the study participants. The proportion wanting no more children increased with age, number of children and level of education among both men and women. We also found that the proportion wanting no more children increased with increasing duration of marriage. Fertility intentions differed by knowledge of HIV status. Both men and women who reported that they were HIV positive, either on ART or not on ART, were twice as likely to report a desire to stop child bearing compared to men and women who reported that they were HIV negative.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image ()
• TIFF
  original image ()

Table 1. Fertility intentions by background characteristics.

doi:10.1371/journal.pone.0051861.t001

Using Fishers Exact Test we found no significant difference between those on ART and those not on ART and therefore combined these two groups.

Table 2 presents unadjusted and adjusted odds ratios (OR) predicting the likelihood of wanting no more children among married men and women. Adjustment for a wide range of possible confounders made little difference. We found that HIV positive individuals were more likely to want no more children compared to their HIV negative counter parts, with adjusted odds ratios of 4.9 (95% CI 2.7–8.9) for men and 7.4 (95% CI 3.2–16.7) for women.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image ()
• TIFF
  original image ()

Table 2. Unadjusted and adjusted odd ratios (OR) predicting men’s and women’s likelihood of wanting no more children by HIV status.

doi:10.1371/journal.pone.0051861.t002

Couples’ HIV Awareness and Fertility Intentions

Table 3 shows fertility intentions by the couples’ HIV status. The expressed desire for both to have no more children was highest 27/38 (71%) among concordant HIV positive couples and lowest 325/1260 (26%) among concordant HIV negative couples, with discordant couples having an intermediate value 33/61 (54%). Among concordant negative couples, husbands and wives were equally likely to want no more children. Among concordant positive and discordant couples, the majority agreed in wanting no more but, in cases of disagreement, it was the husband who was more likely to want no more.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image ()
• TIFF
  original image ()

Table 3. Couples’ fertility intention by knowledge of their HIV status.

doi:10.1371/journal.pone.0051861.t003

For the subsequent analysis couples with at least one member wanting more or undecided were combined with those where both wanted more or were undecided. Logistic regression in Table 4 shows that concordant HIV positive couples were more likely to desire to stop child bearing than were their negative counterparts: odds ratio 6.6 (95% CI 3.0–14.8), increasing to 11.2 (95%CI 4.7–26.9) after adjusting for possible confounders. We also found that discordant couples were also more likely to express the desire to stop child bearing compared to couples who reported being HIV negative or couples who reported that they did not know their HIV status.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image ()
• TIFF
  original image ()

Table 4. Unadjusted and adjusted odds ratios of couples’ fertility intentions by couples’ composite knowledge of their HIV status.

doi:10.1371/journal.pone.0051861.t004

Among those who did not want more children the most important reason given varied. Among HIV negative women who wanted no more children, finances or effects on woman’s health were major foreseen adverse consequences, each reported by 291/351 (83%). Among HIV positive women effects on woman’s health was the major reason, reported by 15/20 (75%) of women on ART and 10/15 (67%) of women not on ART. For HIV negative men, 206/313 (66%) reported finances and 74/313 (24%) the woman’s health. Among HIV positive men on ART 14/33 (42%) reported women’s health and 9/33 (27%) child health. Among HIV positive men not on ART 6/18 (33%) reported woman’s health, 5/18 (28%) finances and 5/18 (28)% child health.

Couples’ HIV Awareness and Contraception

Levels of contraceptive use reported by wives were almost identical among concordant negative, concordant positive and discordant couples (Table 5). Little difference was apparent between concordant couples in method choice, with hormonal methods and condoms equally popular. However, hormonal methods were less commonly used by discordant couples and condoms were more likely to be used. Contraception use did not seem to vary depending on whether it is the woman or the man who wanted more children. In unions where a woman wanted more children and the man wanted to stop 63/143 (44%) women used any modern method of contraception compared to 60/143 (42%) in unions where a woman wanted to stop and a man wanted more children. Logistic regression showed no significant difference in contraceptive use between couples by HIV status. Compared with concordant negative couples, adjusted odds ratio of use of any modern method of contraception in the concordant positive couples was 1.5 (95%CI 0.8–3.3). The odds ratio for discordant couples was 0.6 (95%CI 0.4–1.0) (Table 6). After controlling for HIV status and other background characteristics couples who agreed to stop child bearing were more likely to use any method of contraception (1.7 95%CI 1.2–2.4).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image ()
• TIFF
  original image ()

Table 5. Use of contraception by couples’ knowledge of HIV status.

doi:10.1371/journal.pone.0051861.t005

Download:

• PPT
  PowerPoint slide
• PNG
  larger image ()
• TIFF
  original image ()

Table 6. Logistic regression model presenting the probability of using contraception by couples’ awareness of their HIV status and stated intentions.

doi:10.1371/journal.pone.0051861.t006

Discussion

In this population based cross-sectional study of monogamously married men and women we found that most HIV positive men and women wanted to stop child bearing. 39/56 (70%) of men on ART and 24/36 (69%) of men not on ART desired to stop child bearing. Similar proportions were found among women. These proportions are higher than those found in others studies [22], [23], [24]. Our findings confirm those of other individual level studies: that awareness of HIV positive status is associated with the desire to stop child bearing among married men and women [8], [25]. We do not see much effect of ART on couples’ fertility intentions. This could be because: 1) the ART program in this country was rolled out to rural areas three to four years later than the time ART program was initiated in urban areas [26] and therefore participants on the ART program have not been on the program long enough to fully appreciate the impact of the treatment on health; 2) high levels of stigma surrounding being HIV positive and ART enrolment could not only prevent people form accessing ART treatment but also prevent disclosure and open discussion in relation to sexual and reproductive health [27]; 3) the numbers of those on ART included in our analysis are too small for us detect the influence of ART on fertility intentions;. 4) HIV positive women and men on ART are counselled to avoid pregnancy so they may not want it and participants may have considered it socially desirable not to express desire for children when HIV positive. These expressed differences in fertility desires by couples’ awareness of their status could assist reproductive health services planners in designing programmes that address couples’ aspirations and needs especially when integrating family planning and HIV prevention interventions in the context of ART programmes.

Use of any modern method of contraception was slightly higher in this setting than in similar settings in Malawi [28]. 628/1302 (48%) of HIV concordant negative, 15/29 (52%) of HIV concordant positive, 13/26 (50%) discordant husband positive and 5/15 (44%) discordant wife positive used any method of contraception. Condom use, especially among HIV discordant couples, was lower than that reported in other studies [22]. Condoms have been described as ‘an intruder’ in marriage in rural Malawi and associated with extra-marital sex [29] and therefore couples may find it difficult to introduce condoms within marriage, especially if the HIV positive partner has not disclosed their status to their spouse [27].

There are a number of limitations of our study. Disclosure of HIV status in this setting and in similar settings has been associated with HIV negative results and therefore its influence on fertility intentions may have been on HIV negative couples more than on HIV positive couples or discordant couples [27]. Our analysis has focussed on monogamously married couples and therefore the findings cannot be generally applied to all couples including those in polygamous unions which constitute 20% of all married unions in the study population.

Due to small numbers our analysis has been unable to establish whether in discordant couples the effect of HIV status on fertility intentions differs depending on whether it is the husband or the wife who is positive. Our analysis however is one of the few studies on this subject involving relatively larger numbers of monogamously married couples.

Conclusion

We have demonstrated that awareness of HIV positive status is associated with the desires to stop child bearing and that members of couples who both want no more children are more likely to use contraception – but contraceptive use is quite low. Greater promotion of an environment which facilitates condom use and contraceptive use within marriage in this and similar settings is warranted, such as couples counselling, linking HIV services and family planning, and taking into account couples’ expressed fertility preferences. As more HIV positive individuals and couples get on ART, we recommend a further analysis of the impact of ART on increased fertility desires and fertility. We also recommend interventions that promote couple disclosure of HIV status as this will assist couples in their efforts to fulfil their reproductive rights.