Risky Behaviors among HIV Serodiscordant and Seroconcordant Couples in Yaounde-Cameroon

Introduction: Serodiscordant HIV couples are an important source of HIV spread in sub-Saharan Africa. To limit the transmission of HIV within couples, UNAIDS has defined many strategies, including unconditional treatment of the infected partner, systematic condom use, assisted reproduction, regular screening of the negative partner. In Cameroon, there is little information on the follow up of HIV serodiscordant and seroconcordant couples. The purpose of this work was to describe risky behaviors within serodiscordant couples. Method: This cohort study was conducted in five health facilities. The data were collected using a survey form and the blood collected was used to test consenting participants for HIV, HbsAg, HCV, syphilis, and chlamydia. Data were analyzed using the GraphPad Prism 6 software. Results: Eighteen seroconcordant and fifty tree serodiscordant couples agreed to participate in the study, where women were predominantly infected (56.66%). Serodiscordance was known for 64 months on average. Among serodiscordant couples, 86.79% infected partners were on ARV treatment, among them 10/46 (15.22%) stopped treatment. Sexual relationship existed in 50/53 couples, with non-systematic use of condom in 44 couples (83.01%). Extramarital relationships were reported in 10 of the 23 serodiscordant couples (56.60%) with libido disorders. HBsAg was most present in HIV infected women (11.32 %) while, HIV-uninfected men were most infected with viral hepatitis C (7. 55 %). Conclusion: This work has highlighted many risky behaviors in serodiscordant couples. It remains essential to provide continuous counselling to serodiscordant couples and to associate systematic STI screening in all cases of HIV/ AIDS serodiscordance.


INTRODUCTION
Are and management of Human Immunodeficiency Virus/ Acquired Immunodeficiency Syndrome (HIV/AIDS) pandemic remains a challenge for low-income countries where sexually transmitted diseases are predominantly transmitted in stable couples in which one partner is or is not receiving care for the infection [1]. In sub-saharan Africa, HIV serodiscordance affects one to two-thirds of couples living with HIV [2,3]. In Zambia and Rwanda, for example, new heterosexual infections in adults from serodiscordant couples range from 55.1 to 92.7% [4]. In 2/6 serodiscordant couples, HIV-negative partners are at a very high risk of becoming infected [5]. HIV acquisition can be fueled by sexually transmitted infections (STIs) that cause ulceration or not. STIs greatly increase the risk of HIV infection by increasing the susceptibility of the exposed person, as well as the infectivity of an HIV-positive person [6]. The presence of an STI in an exposed person increases their susceptibility to HIV, through various mechanisms such as the presence of genital ulcers or local inflammation. For example, the presence of syphilis in an HIV infected person increases the amount of virus in the genital secretions and therefore the risk of contamination of its partners by HIV. This explains the recommendations to offer STI testing before any HIV diagnosis and vice versa [7].
Conscious of the implication of serodiscordance in the occurrence of new cases and the potentiation of the pandemic to HIV / AIDS, WHO through its UNAIDS program has defined strategies for HIV prevention among serodiscordant couples. These strategies include: placing the infected partner on antiretroviral therapy (ARV) unconditionally, using condoms consistently, using ARVs post-exposure and pre-exposure, screening the negative partner on a regular basis, as well as adequate moral support to helping couples manage HIV serodiscordance [8]. The implementation of these strategies is a struggle in Africa because of the lack of a real policy of serodiscordant couple's management according to the contexts. Thus, many risky behaviors (non-systematic condom usage, extra-marital sex, treatment interruption and co-infections with other STDs) have been reported in several studies including those of Ruzagira et al. 2011 in Uganda [9], Gitonga et al., in kenya in 2012 [10], Guira et al., in 2012 in Burkina Faso and many other authors [11].
In Cameroon, there is no real policy on the management of serodiscordant couples, therefore, the notion of serodiscordance in general, the national prevalence, the profile of STIs, the challenges faced by these couples, sexual behaviors within couples are poorly known and the documentation almost nonexistent. Walque in 2006, is one of the few authors who took an interest in serodiscordant couples in Cameroon where he described sexual behaviors in this group [2].
Do they receive the appropriate support to cope with HIV serodiscordance? Do they practice their sexuality in strict compliance with HIV prevention recommendations at the individual and couple level as envisioned by WHO? The purpose of this work was to describe, 13 years after the study of Walque et al., the risky behaviors in serodiscordant couples that may promote HIV/AIDS transmission to the seronegative partner.

MATERIALS AND METHODS
This was a cohort study conducted from September 2016 to December 2017 in five hospitals (District Hospital of Efoulan, Nkoldongo, Biyem-Assi, cité-verte, Obala and CMA of Nkomo).
Serodiscordant couples consulting in one of these hospitals were our study population, including seroconcordant couples as controls. The term couple refers to two persons living in a household and related by marriage or concubinage for a period of at least twelve months and having sex together. The medical records of HIV positive patients were consulted to identify serodiscordant couples, after which they were contacted by telephone. For those who consented, an appointment was made. The data were collected using a questionnaire offered to each partner separately or together. This questionnaire collected sociodemographic data, therapeutic follow-up, knowledge and attitudes regarding HIV/AIDS. Subsequently, each of the partners was blood sampled and the presumed HIV-negative partners were tested for HIV following the national HIV / AIDS screening algorithm: a first rapid test using Determine ® HIV1/2 (Abbott Diagnostic Division, Hoofddorp, The Netherlands) and a second confirmatory test with KHB ® HIV (Shanghai Kehua Bioengineering Co., Ltd., China). In case of discordant results to these two tests, the sample was re-analysed by ELISA. The presence of HBsAg, anti-HCV antibodies were determined by the LabACON ® kit (Hangzhou Biotest Biotech Co., Ltd., China), anti-treponemal antibodies were searched using two kits, TPHA Biolabo and RPR (Biolabo SA, 02160, Maizy France) and anti-Chlamydia trachomatis antigens were investigated by the Immunocomb ® Chlamydia trachomatis IgG kit (Orgenics, Israel). The data were entered and analyzed using the GraphPad Prism Version 6 software. The χ2 or Fischer test was used to compare the qualitative variables and the Student's test to compare the means with a significance level set at 5 %. The study was approved by the National Research Ethics Committee for Human Health (N° 2016/04/ 758/ CE/CNERSH/SP). Before conducting the study, written informed consent was obtained and all identifying information (names, initials.) were then omitted to ensure the confidentiality.

Sociodemographic characteristics of couples
A total of 266 couples were received. Among them, those who accepted to participate were grouped as follows: 60 / 192 couples (31.25%) were serodiscordant and 19 / 74 (25.67%) seroconcordant. Eight couples were excluded, 7 serodiscordant couples (three for subsequent seroconversion of negative partners and 04 for insufficient data), and one seroconcordant couple (for insufficient data). All couples were heterosexual and monogamous. A proportion of 56.60% of women (30) versus 43.39% of men (23) were infected (p = 0.1739). In addition, 60.38% (32) of the couples were married, while 39.62% (21) lived in concubinage for an average of 03 years and 03 months (ranging from one year to 17 years). Serodiscordance was known for 5 years and 2 months on average (min-max from 3 months to 15 years). All respondents reached at least primary level of education. They were more men (8/53) with a university level than women (5/53). The age of women ranged from 17 to 59 years with an average of 32.81; while that of the men was of 22 and 67 years and an average of 41.26. Although 37out of 53 (69.81%) couples wanted children, 18.91% (07/37) did not have children. Nevertheless, 9 couples (16.98%) did not want to have children anymore (Table 1).

ART observance
We observed in this study that seroconcordant couples were more observant for their ARVs treatment compared to serodiscordant couples (P = 0.0225). All seroconcordant couples have not uninterrupted their treatment, while 21.74% (10/46)   (Table 2).

Sexual behaviors within couples
Of the 53 serodiscordant couples surveyed, sexual intercourse and abstinence were observed in 50/53 (94.34%) and 03/53 (5.66%) couples, respectively. Altered sexual desire or maintenance of sexual desire was equitably distributed (p = 0.4201) at 43.39% (23/53) and 56.60% (30/53) between those who reported sexual desire disorders compare to those with no libido disorders in sero-discordant couples. Among couples who reported sexual desire impairment, 13.04% of couples (03/23) opted for abstinence while sexual intercourse became increasingly rare, ranging from one to two times within three months in 17.39% of them (04/23) and one time in six months for 4.34% (1/23) of couples. Although libido disorders were reported in 50% (09/18) of seroconcordant couples, the frequency of sexual intercourse was on average once per week.  Table 3).

Prevalence of STIs (viral hepatitis B and C, syphilis and chlamydia)
In general, none of the STIs was more present in HIV infected partners than HIV non infected partners (p = 0.559) Table 5. In HIV-infected women, 13.33% (4/30) were HBV-positive (HBsAg) versus only one among HIV-negative women (1/23(4.3%)) (p = 0.039). HIV-positive women were 3.25 times more likely to be infected with HBV than HIV-negative women. The distribution of the other three STIs (HCV, syphilis and chlamydia) was similar between HIV infected and HIV non infecetd women. In contrast, more HIV-negative men (13.33%, 4/30) were infected with viral hepatitis C (P = 0.039) compared to HIV infected men (4.3%, 1/23). Syphilis was more present in HIV negative men 3/30 (10 %) than HIV positive men (1/23; 4.3%) without significant difference (p= 0.40). Only one case of syphilis and one of chlamydia were reported in HIV positive women and HIV negative men respectively.   A case of HBV and HVC co-infection was observed in an HIV infected man.

DISCUSSION
Management and care for HIV serodiscordant couples require careful monitoring of both partners to limit the risk of HIV transmission as well as other STIs [1]. However, the low study participation rate estimated at 68.75% and 74.33% respectively among serodiscordant and sero-concordant couples with HIV, may translate little sensitization of couples for their contribution in care in Cameroon. Greater participation in the study would have made it possible to better appreciate the realities or experiences of HIV serodiscordant couples in Yaounde.
This study has identified some risky sexual practices like poor condom usage, poor therapeutic compliance, and high prevalence of four STIs among serodiscordant couples compared to seroconcordant HIV couples. All HIV-negative partners participating in this study repeated their test at least twice from first testing. In serodiscordant couple, female partners were more infected with HIV than men partners but without the significant difference. This trend was observed in several studies including those of Walque, 2006, Colman-Sarfo et al., 2017 [2,12]. This could be explained by the fact that women are physiologically more susceptible to HIV than men [13]. It should also be remembered that women can, in many cases, be forced to have sex [2].
HIV seroconcordant couples were more compliant to ARV treatment compared to serodiscordant couples where 18.86% (10) of the HIV positive partners discontinued treatment. The main reasons for discontinuation of ARVs were the stock-out in 4 cases (40%) and in 6 cases (60%) discontinuation was due to mental disorders resulting from the non-acceptance of HIV status.
This risky behavior (the interruption of treatment) increases the chance of transmission to the negative partner as stopping the treatment leads to the increase of viral load, therefore the risk of transmission [14]. Previous work done by Troussier and Tourette-Jurgis in 2006 in France showed that the prevalence of sexual desire disorder was twice as high among serodiscordant couples compared with seroconcordant couples [15]. Similarly, Guira et al (2013) in Burkina Faso observed a very high rate of sexual desire disorders in 97.7% discordant couples compared to 56.60% in our study. Sexual desire disorder is justified in all these works mainly by the fear of infection of the negative partner [11]. Specifically, the low frequency of sexual desire disorders in the present study compared to the other two studies could be justified by the increase in the number of CTAs (Approved Treatment Centers), UPECs (HIV care and Management Unit) and their decentralization up to at the lowest level of the health pyramid that provide psychological support to the patient living with HIV [16].
Extramarital relationships (by the seronegative partner), the major consequences of sexual desire disorders in serodiscordant couples, were observed in 18.86% of couples, which represent a risk of exogenous acquisition or transmission not only of HIV, but also of STIs within couples. This result is close to that found in two studies, one in New York and the other in Burkina Faso with respectively 18% and 25% of couples with extramarital relationships [17,11]. It should be noted that this behavior is observed most in the male partner than the female partner in all these studies including the work of Walque in 2006 in 5 African countries (including Cameroon). This could be justified by the fact that in the African context, in almost all cultures, traditional customs encourage men to have multiple partners [18].

5/6
no link was found between libido disorders and extramarital relationships, a correlation has been established in our study between extramarital relationship and sexual desire disorder [17,11]. Serodiscordant couples were twice as likely to have extramarital relationships when libido disorders were reported compared to seroconcordant couples. The frequency of 18.86% of extramarital relationships recorded in this work may not reflect the true trend of this risk behavior among couples because one of the weaknesses of this study was the fact of having interviewed 29 couples (54. 72%) together. In fact, these couples claimed to have nothing to hide and demanded to participate together in the study. All the cases of extramarital affairs reported came from the 45.28% (24/53) couples interviewed separately. These findings suggest an urgent need to strengthen couples' counselling on a psychological level in order to help them manage serodiscordance [19]. The male condom was more frequently used in seroconcordant couples compared to serodiscordant couples. This can be explained by the fact that both partners who are infected regularly receive advice on infection management, so they can easily apply HIV / AIDS prevention strategies compared to serodiscordant couples where only one partner is involved and benefits from the above mentioned advices. Studies have shown that the correct use of condoms reduces the risk of HIV transmission by 80% [20]. Condom use in a non-systematic way or never as declared by certain serodiscordant couples (79.24%) represents a major endogenous risk for the infection of the seronegative partner. This proportion of 79.24% is higher than that of Guira et al [11] where 59.5% of couples did not use condoms consistently, as well as those of Colman-Sarfo et al [12] where 40% did not use condoms. These findings suggest the need to define specific control strategies for this group in order to limit the spread of HIV infection because even in the case of undetectable viral load of the HIV virus, UNAIDS recommends the consistent use of condoms in serodiscordant couples [1]. The main reason for not using condom was the need to procreate in 69.81% of cases, and this behavior is observed in couples with no children as well as those with children. Moreover, in the African sociocultural context, procreation is a social pressure for the life of a couple [21]. Low medical assistance is therefore unfavorable to procreation without risk of HIV transmission, including STIs to seronegative partners in serodiscordant couples [22,23]. The distribution of STIs in the HIV positive and HIV negative partners showed that, apart from HbsAg which was more present in the HIV-positive partner, the rest of the three STIs were more prevalent in the HIV-negative partner, which is surprising when we know that HIV infection promotes the occurrence of other diseases in the sense that it decreases the immunity [8]. This observation may be justified by the fact that HIV-negative partners may be more susceptible to extramarital relationships, which predisposes them to exogenous exposure to STIs [24]. HIV / AIDS influences the natural history of HBV [25]. In this work, HIV-positive women were more infected with viral hepatitis B than HIV-negative women with a risk estimated at 3.25 times to contract viral hepatitis B when the woman was HIV positive. This co-infection, estimated at 13.33% among women in this work, is greater than 1% found in the work of Dovonou et al., in 2015 in Benin [26]. This difference in outcome could be justified by the difference in distribution of viral hepatitis B according to the regions in sub-Saharan Africa [27]. Although HIV influences the natural history of viral hepatitis C as well [28], HIV-negative men were more infected with viral hepatitis C than HIV positive men. This observation could be explained by the fact that HIV-negative men would engage in extramarital affairs and therefore expose themselves to the risk of acquisition of STDs than HIV infected with reduced libido.

CONCLUSION
In light of the results of this work, which revealed many risky behaviors among serodiscordant couples, it remains essential to define specific HIV/AIDS prevention strategies among serodiscordant couples as a means of limiting new infections. Targeted counselling should be developed to promote assisted and safe reproductive methods in discordant couples who need to procreate. With regard to STIs, it will be necessary to associate their systematic screening not only in serodiscordant partner who is invited to be tested for HIV infection on a regular basis, but also in the general population.