Increasing access to HAART has extended life and reduced morbidity and mortality among HIV-infected individuals [1,2]. However, as HAART improves physical health, HIV-infected individuals may be more likely to engage in unsafe sex (i.e. risk compensation or behavioral disinhibition) [3–6]. On the other hand, increased access to clinical care via HIV prevention counseling and receiving HAART may encourage reductions in sexual risk behavior. As HAART becomes more readily available in resource-limited settings, the relationship between treatment and sexual behavior also becomes more important [7,8]. A key component of the recently proposed HIV prevention strategy of ‘Universal Test and Treat,’ in which everyone diagnosed with HIV infection is promptly initiated on HAART to decrease HIV transmission by reducing population viral load, partially depends on sexual risk behavior not increasing once individuals are on HAART [9,10].
South Africa has an adult HIV prevalence of 18.1% and has the largest HIV epidemic in the world with 5.7 million HIV-infected individuals . Since 2005, a growing number of HIV-infected South Africans have gained access to HAART so that by 2007, a third of South Africans in need of treatment were receiving HAART . In South Africa, the high prevalence of HIV is likely fueled by a range of high-risk sexual behaviors, including inconsistent condom use, early age at sexual debut, and multiple sex partners [13–17].
Secondary prevention efforts that aim to decrease the risk of HIV transmission from already infected individuals through behavioral and biomedical interventions have yet to be widely implemented in southern Africa [7,18]. There is some evidence that HAART may not be associated with an increase in sexual risk-taking behavior in sub-Saharan Africa , though much of these data have included limited patient follow-up, were cross-sectional, or were taken exclusively from high-risk groups [19–26]. With increasing efforts to expand treatment access and secondary prevention in the region, quantifying and comparing potential shifts in sexual behaviors over time among populations of HAART initiators remains an urgent priority.
We examine the relationship between HAART status and being sexually active, having unprotected sex, and having more than one sex partner before and after HAART initiation, overall and stratified by sex, among an operational cohort of South African men and women enrolled in urban and rural primary care programs.
Setting and participants
Between June 2003 and January 2010, HIV-infected patients were recruited to participate in an operational cohort study at HIV primary care and support clinics in urban and rural South Africa. Eligible participants were at least 18 years old, HIV-infected, and had given written informed consent. The urban clinic was at the Perinatal HIV Research Unit (PHRU) at Chris Hani Baragwanath Hospital in Soweto, and the rural at Tintswalo Hospital in Mpumalanga Province. Participants were either self-referred for HIV care or were referred from mother-to-child transmission prevention programs, hospital wards, or HIV counseling and testing programs.
Clinical care, delivered primarily by nurses with physician oversight, included the following HIV-related clinical services: symptom screening for sexually transmitted infections (STIs) and tuberculosis , prophylactic treatment for opportunistic infections, treatment for STIs and HIV-related opportunistic infections, annual cervical smears , and female family planning methods. Clinical and research study personnel underwent training sessions that involved rehearsing the survey questions, training in nonjudgmental survey delivery, and instruction in prevention counseling. Since 2004 at PHRU and 2005 at Tintswalo, participants were referred for or offered HAART when they were diagnosed with WHO stage 4 disease or when their CD4 cell counts fell below 200 cells/μl per South African guidelines [29,30]. Patients were enrolled in care prior to treatment initiation and gave written informed consent prior to data being collected. At the time of HAART initiation, all patients were offered adherence counseling. All care was provided free of charge. The Ethics Committee at the University of the Witwatersrand approved this study.
At baseline, participants were interviewed using a standardized questionnaire that collected demographic, socioeconomic, behavioral, and health history information. Participants were interviewed using a similar instrument at data collection visits scheduled 4–7 months apart, but could visit the clinic at any time for care. Laboratory data included CD4 cell count at baseline and every 6 months. For the current study, we assessed sociodemographic variables: age, employment, urban or rural residential status, marital status, number of people living in household, and current contraception status (for women); and behavioral variables: sexual activity, alcohol use, needing assistance in completing daily activities (as a measure of general health status), and disclosure of HIV status. At baseline and study follow-up, participants were asked about their sexual behaviors, including whether or not they had a regular partner, whether they had a casual sex partner, and the frequency of condom use with both regular and casual partners (if they had one). Recall for these questions was since the last clinic visit. Information on condom use for each partner type was collected using categorical variables, and consistent condom use was defined as self-report of always using a condom with a sex partner. These measures of sexual behavior have been employed in this setting .
We assessed three measures of self-reported sexual behavior as outcomes in the current study: being sexually active; having unprotected sex with a regular and/or casual partner; and having more than one sex partner. As not all participants reported a regular or casual sex partner, we first examined factors associated with being sexually active. The outcomes of having unprotected sex and having more than one sex partner were analyzed only among those who reported being sexually active. We restricted our analyses to CD4 cell counts taken within 90 days of a study visit. Data were analyzed according to an intent-to-treat analysis, in which once participants were initiated on HAART, they were considered HAART-experienced at all subsequent clinic visits. To assess the effect of HAART initiation on potential changes in sexual risk behavior, we compared sexual behavior during pre-HAART visits to behavior reported after HAART initiation. After introducing HAART status –the primary predictor variable – into the models, each covariate was introduced into the three models to assess potential confounding based either on a change of at least 10% of the nonlog transformed beta coefficient of independent risk factors, or a priori confounders based on a review of earlier studies. The three sexual behavior outcomes were assessed as time-varying outcomes. CD4 cell count and HAART status were adjusted for as time-varying covariates. The baseline status of all other sociodemographic and behavioral covariates was also adjusted for in the models. We controlled for time-dependent changes in risk behavior by adjusting for years since enrollment in care via calendar year category (2002–2004, 2004–2006, and 2006–2009) . We employed generalized estimating equations (GEEs) with a logit link and exchangeable correlation structure to assess the impact of HAART on sexual risk behavior utilizing PROC GENMOD in SAS . To further examine within-participant variation (i.e. treatment effect) while still controlling for stable participant characteristics, we also employed fixed effects regression methods utilizing PROC LOGISTIC in SAS ; however, this analysis was broadly similar to the GEE analysis and is, therefore, not presented here. We examined three multivariable models of sexual behavior using GEE, namely, being sexually active (Model I), unprotected sex (Model II), and having more than one sex partner (Model III). In a sensitivity analysis, these three outcomes were re-examined by restricting analyses to those participants who initiated HAART during follow-up. We also assessed the impact of CD4 cell count on sexual behavior through examining clinic visits for which CD4 cell count data were available. All analyses used STATA (version 10.0; STATACORP, College Station, Texas, USA) and SAS (version 9.2; SAS Institute, Cary, North Carolina, USA) software.
Characteristics of participants
Between 2003 and 2009, there were a total of 6263 participants, of whom 4719 (75.3%) were women, accounting for a total of 19703 clinic visits (4479 for men and 15224 for women). During the study period, 2332 participants (37.2%) initiated HAART, of whom 71.7% were women. Table 1 presents the characteristics of all 6263 participants, stratified by sex and treatment status. The median age at study enrollment was 33.8 years [interquartile range (IQR): 29.1–40.0] and the median CD4 cell count at study enrolment was 253 cells/μl (IQR: 110–440). The urban cohort was larger (65.3%), and most participants were unemployed (76.4%) and never married (53.8%). At baseline, most (87.3%) had disclosed their HIV status, and 33.2% reported alcohol use. About a third enrolled into care between 2002 and 2004, 2005 and 2006, and 2006 and 2009, respectively. The mean number of study visits was 3.22 (SD 2.45) per person, with a mean number of 2.64 (SD 1.59) post-HAART visits and 2.59 (SD 2.15) pre-HAART visits per person. At study enrollment, compared to those who did not initiate HAART, participants who did start HAART during the study were less likely to be women (71.7 vs. 77.5%; P < 0.0001), to be urban residents (42.6 vs. 78.7%; P < 0.0001), to be employed (18.3 vs. 26.8%, P < 0.0001), and to be never married (48.6 vs. 56.9%; P < 0.0001), and were more likely to be older (age >40 years: 32.6 vs. 20.5%; P < 0.0001).
Sexual behavior and HAART
Table 2 presents the frequency of being sexually active (Model I), reporting unprotected sex (Model II), and having more than one sex partner (Model III) at visits before and after HAART initiation. Out of a total of 9568 pre-HAART and 3496 post-HAART visits, participants were more likely to be sexually active before HAART compared to after HAART initiation (70.7 vs. 56.6%; P < 0.0001). By partner type, participants with both a primary and casual partner, a primary partner only, and a casual partner only, reported a higher frequency of being sexually active before HAART initiation compared to after HAART initiation (P < 0.0001). Both men and women were more likely to report being sexually active before HAART initiation (P < 0.0001). In regard to unprotected sex, of 13064 visits at which sexual activity was reported, participants reported unprotected sex at 1968 pre-HAART (20.6%) visits compared to 346 post-HAART (9.9%) visits (P < 0.0001). By partner type, participants with both a primary and casual partner and only a primary partner reported a higher frequency of unprotected sex before HAART initiation than after HAART initiation (P < 0.05). Both men and women reported unprotected sex at a higher frequency before HAART initiation (P < 0.0001). Of 600 visits during which having more than one sex partner was reported, participants reported having more than one sex partner at 542 pre-HAART (5.7%) visits compared to 58 post-HAART (1.7%) visits (P < 0.0001), a difference that held across sexes. We also examined these three sexual behavior outcomes only among the subset of participants who initiated HAART who contributed a total of 9032 visits. Similar to the analyses above, participants reported a higher frequency of being sexually active, engaging in unprotected sex, and having more than one sex partner at pre-HAART visits compared to post-HAART visits (P < 0.0001).
Adjusted analyses of HAART and being sexually active (Model I) overall and stratified by sex
Overall, the likelihood of being sexually active decreased after HAART initiation [adjusted odds ratio (AOR): 0.86 (95% confidence interval, CI: 0.78–0.95); P = 0.0036], and this decrease was greater for women [AOR: 0.40 (95% CI: 0.35–0.46); P < 0.0001] than for men [AOR: 0.90 (95% CI: 0.74–1.08); P = 0.2745; see Table 3]. Other independent predictors of decreased sexual activity included female sex [AOR: 0.76 (95% CI: 0.66–0.87); P = 0.0001], older age [AOR >40 years: 0.31 (95% CI: 0.27–0.36); P < 0.0001], and needing assistance with daily activities [AOR: 0.59 (95% CI: 0.50–0.71); P < 0.0001]; and of increased sexual activity included urban residence [AOR: 1.97 (95% CI: 1.71–2.26); P < 0.0001] and married/living together [AOR: 4.42 (95% CI: 3.85–5.08); P < 0.0001]. Though not significant predictors overall, unemployment for men and alcohol use, injectable hormone-based and pill-based contraception, and enrollment in care since 2006 for women were significantly associated with being sexually active.
Adjusted analyses of HAART and unprotected sex (Model II) overall and stratified by sex
Overall unprotected sex decreased after HAART initiation [AOR: 0.40 (95% CI: 0.34–0.46); P < 0.0001], and this decrease was greater for men [AOR: 0.33 (95% CI: 0.24–0.44); P < 0.0001] than for women [AOR: 0.71 (95% CI: 0.59–0.86); P = 0.0005; see Table 4]. Overall other independent predictors of increased unprotected sex were female sex [AOR: 1.63 (95% CI: 1.37–1.95); P < 0.0001], marred/living together [AOR: 1.27 (95% CI: 1.10–1.46); P = 0.0010], STI symptoms [AOR: 1.35 (95% CI: 1.20–1.53); P < 0.0001], and older age [AOR for >40 years: 1.25 (95% CI: 1.03–1.53); P = 0.0229]; and of decreased unprotected sex were urban residence [AOR: 0.71 (95% CI: 0.58–0.85); P = 0.0004] and enrollment in care since 2006 [AOR: 0.71 (95% CI: 0.59–0.85); P = 0.0003]. Though not significant predictors overall, needing assistance with daily activities for men and unemployment, hormone-based contraception, and disclosure of HIV status for women were significantly associated with unprotected sex.
Adjusted analyses of HAART and having more than one sex partner (Model III) overall and stratified by sex
Overall the likelihood of having more than one sex partner decreased after HAART initiation [AOR: 0.20 (95% CI: 0.14–0.29); P < 0.0001], and this decrease was greater for women [AOR: 0.09 (95% CI: 0.05–0.17); P < 0.0001] than for men [AOR: 0.35 (95% CI: 0.23–0.53); P < 0.0001; see Table 5]. Other independent predictors associated with a decreased odds of having more than one sex partner were female sex [0.45 (95% CI: 0.36–0.56); P < 0.0001], married/living together [AOR: 0.53 (0.41–0.67); P < 0.0001], and enrollment in care since 2005 [AOR for 2005–2006: 0.20 (95% CI: 0.15–0.27); P < 0.0001 and AOR for 2006–2009: 0.19 (95% CI: 0.13–0.29); P < 0.001]. Though not significant predictors overall, urban residential status and disclosure of HIV status for men and older age (AOR for >40 years) for women were significantly associated with having more than one sex partner.
Restricting analyses to the subset of 2307 participants who initiated HAART during follow-up time did not alter the overall association between HAART and being sexually active [AOR: 0.87 (95% CI: 0.77–0.98); P = 0.0238], unprotected sex [AOR: 0.43 (95% CI: 0.36–0.52); P < 0.0001], and having more than one sex partner [AOR: 0.17 (95% CI: 0.12–0.24); P < 0.0001]. Additionally, to examine the impact of behavior change following HAART initiation that was independent of CD4 cell count, we examined the three behavioral outcomes adjusting for CD4 cell count in the 5704 visits with CD4 cell counts (65.2% were pre-HAART and 34.8% were post-HAART); outcomes were consistent with the earlier analyses for the association between HAART and being sexually active [AOR: 0.78 (95% CI: 0.67–0.93); P = 0.0047], unprotected sex [AOR: 0.32 (95% CI: 0.23–0.44); P < 0.0001], and having more than one sex partner [AOR: 0.25 (95% CI: 0.13–0.50); P < 0.0001]. We also assessed effect modification by strata of CD4 cell count based on cut-offs that have been considered for treatment initiation (<200 cells/μl, 200–350 cells/μl, vs. >500 cells/μl) and these stratum-specific AORs were not significantly different from each other, suggesting that sexual risk behavior did not differ by CD4 cell count strata.
In this population of HIV-infected South African men and women enrolled in urban and rural primary care programs, we documented a strong and sustained reduction in sexual risk behaviors following HAART initiation. Among both men and women in these two cohorts, we observed reductions in those reporting sexual activity, unprotected sex, and more than one sex partner, suggesting that HAART in this population is not associated with risk compensation. Furthermore, these findings were independent of CD4 cell count. To our knowledge, this is one of the first studies to report on sexual risk among a longitudinal HIV-infected cohort spanning nearly a decade of observation before and after HAART initiation in a resource-limited setting. These findings add to our understanding of how HAART could further assist in reducing the secondary transmission of HIV.
There has been a paucity of longitudinal data examining the relationship between initiating HAART and sexual risk among HIV-infected Africans. Our findings from a large clinic-based sample of men and women provide greater generalizability than earlier studies from the region. Recent data from HIV-infected female sex workers in Kenya found that unprotected sex but not the number of sex partners decreased following HAART initiation . Also, recent national survey data from Cameroon showed that HAART was associated with consistent condom use [25,26]. South African and Kenyan cohorts with 12 months of HAART follow-up have also documented a decrease in the frequency of unprotected sex [34,35]. A clinic-based cohort from Uganda found that though sexual activity increased following HAART, condom use increased and the number of sex partners decreased . An earlier Ugandan cohort with up to 6 months of HAART also documented reductions in risky sexual behavior . Recent cross-sectional studies from African settings have also suggested that HAART was not associated with increased sexual risk behaviors [20–22,37]. In contrast, a study from Cote d'Ivoire found that unprotected sex significantly increased during the first 6 months of treatment . Differing results in these settings highlight the complexity of measuring the association between HAART and sexual behavior and variations in study populations and treatment program characteristics. As HAART continues to be rolled-out in resource-limited settings and an increasing number of individuals are on treatment for an extended period of time, further quantitative and qualitative research should examine whether initial decreases in sexual risk behavior are sustained and reasons for this. While epidemiological studies can examine whether individuals on treatment engage in sexual risk behaviors, qualitative and biological data are needed to understand specific reasons for reductions in sexual risk behaviors. We unfortunately have no evidence for reasons for our findings; however, apart from counseling prior to HAART initiation, other reasons may include treatment optimism and metabolic effects of HAART.
Sexual activity was reported at two-thirds of study visits overall and decreased following HAART initiation. Data from Uganda and Cote d'Ivoire suggest that a substantial number of patients on HAART do not engage in sexual activity [21,39]. In the current study, despite the finding that HAART was not associated with an increase in unprotected sex, many participants still engaged in unprotected intercourse and hence, the potential risk for HIV transmission persists. Earlier data from South Africa have documented high levels of unprotected sex (i.e. 30–50%) among HIV-infected individuals . It is important that HIV clinical care programs continue to emphasize consistent condom use even after initiating therapy.
Overall urban residents were less likely to report unprotected sex or multiple sex partners. It is possible that urban residents have greater access to prevention messages. Participants who had disclosed their HIV status were less likely to engage in sexual risk behaviors. HIV disclosure was very high in this population (>80%) at clinic enrollment and may suggest that many patients could negotiate condom use and discuss sensitive issues such as sex and relationships with their partners . These findings are similar to an earlier cross-sectional analysis we undertook prior to the wider roll-out of HAART . Additionally, unprotected sex was more frequent among couples who were married or living together, which is in accordance with earlier data from Africa . Data from Uganda among discordant couples showed that HAART was associated with reduced HIV transmission risk and sexual risk behavior decreased following enrollment . Due to the fact that HIV discordance is common within African couples, couple-based prevention in care settings remains important [39,42,43].
The current study included only patients enrolled in primary care programs in which they received ongoing counseling and prevention messages, and hence patients may have been less likely to report risky sexual behaviors. However, study staff received training to ask questions in a nonjudgmental manner to minimize social desirability. Also, due to the comprehensive care and prevention package, it is not possible to disentangle the relative effects of prevention activities, counseling, and condom provision from the effects of providing HAART. It is possible that the documented reductions in sexual risk behavior may have occurred due to the comprehensive nature of care as opposed to HAART alone. Although the current open cohort spanned an 8-year period, participants had two to three visits after initiating HAART, and further longitudinal follow-up will be needed to examine longer-term sexual risk behaviors on HAART. Additional variables that are important determinants of sexual risk behavior that we were unable to assess include desire for children, involvement in commercial sex, drug use, and primary partner HIV status. The current study used a 6-month recall period for sexual risk behaviors, but a shorter period may have been warranted to assess specific sexual behaviors such as condom use. We are unable to comment on reasons for losses to follow-up as this information was not systematically collected, which could have biased these findings. Although the current study utilized both GEE and fixed effects models, bias induced by missing data during follow-up could have occurred . Though patients who did not initiate HAART varied on sociodemographic characteristics, we adjusted for these variables in our multivariable analysis and also conducted further sensitivity analyses excluding patients who did not initiate HAART. To elucidate the impact of HAART on within participant sexual behavior, we also conducted further analyses using fixed effects modeling, and those findings were consistent with the analyses presented here.
Strengths of the current study include prospective data collection before and after HAART initiation. Participants received the same clinic-based HIV prevention and counseling services before and after HAART initiation, reducing potential information bias. A large sample size allowed us to examine a range of confounding variables in our models. We also assessed sexual behavior using a variety of relevant outcome measures, which demonstrated consistent changes toward lower levels of risk taking. Earlier studies in this region have often had short study follow-up or limited generalizability due to being conducted among high-risk subpopulations. The current study can be generalized to HIV-infected patients receiving primary care in both urban and rural settings in sub-Saharan Africa where HAART is increasingly available.
To develop effective secondary prevention interventions in African settings in the era of expanding access to treatment, a broad evidence base about patterns of sexual risk behaviors among HIV-infected individuals before and after initiating HAART is necessary. The current study documents consistent behavioral risk reduction following HAART and provides evidence for the use of HAART as a means of HIV prevention. The public health benefits through the use of HAART for prevention depend not only on the number of individuals treated, but also that treated individuals do not massively increase sexual risk-taking behaviors [45,46]. More interventions are needed to promptly initiate HIV-infected individuals on treatment in resource-limited settings and to consider providing HAART to individuals who currently do not meet treatment guidelines as they may be at considerable risk for transmitting HIV. Integrating access to prompt treatment with appropriate counseling and prevention programs could reduce HIV transmission in the hyperendemic settings of southern Africa.
The authors would like to thank Ms Nkeko Tshabangu for clinical care and training; Mr Tebogo Modisenyane and Mr Yudesh Baliram for data management; and Drs Adele Heyer, Theuns Kotzee, Kgotso Lentle, and Mpolokeng Melamu for clinical oversight.
Patient care is funded by a PEPFAR grant through USAID South Africa (674-A-00–08-00009-00). NIH grants HL090312 and AI048526 supported the analysis. E.M. received research training through Fogarty International Center Grants (2RTW007370/3). The opinions herein do not necessarily reflect those of USAID or the US Government.
K.K.V., N.M., and G.dB. with assistance from M.L. designed the study and wrote the manuscript. K.K.V. with E.M. and E.W.T. did the analyses. L.M., N.T., N.M., M.M. and P.P. provided patient care and oversaw data collection. P.P. and G.E.G. provided oversight on analysis and manuscript writing.
The authors have no conflicts of interest.
1. Palella F, Delaney KM, Moorman AC, Loveless MO, Fuhrer J, Satten GA, et al
. Declining morbidity and mortality among patients with advanced human immunodeficiency virus infection. HIV
Outpatient Study Investigators. N Engl J Med 1998; 338:853–860.
2. Kumarasamy N, Solomon S, Chaguturu SK, Cecelia AJ, Vallabhaneni S, Flanigan TP, Mayer KH. The changing natural history of HIV
disease: before and after the introduction of generic antiretroviral therapy
in southern India. Clin Infect Dis 2005; 41:1525–1528.
3. Dukers N, Goudsmit J, de Wit JB, Prins M, Weverling GJ, Coutinho RA. Sexual risk behaviour relates to the virological and immunological improvements during highly active antiretroviral therapy
-1 infection. AIDS
4. Kelly J, Otto-Salaj LL, Sikkema KJ, Pinkerton SD, Bloom FR. Implications of HIV
treatment advances for behavioral research on AIDS
: protease inhibitors and new challenges in HIV
secondary prevention. Health Psychol 1998; 17:310–319.
5. Stolte I, Dukers NH, Geskus RB, Coutinho RA, de Wit JB. Homosexual men change to risky sex when perceiving less threat of HIV
since availability of highly active antiretroviral therapy
: a longitudinal study. AIDS
6. Ostrow D, Fox KJ, Chmiel JS, Silvestre A, Visscher BR, Vanable PA, et al
. Attitudes towards highly active antiretroviral therapy
are associated with sexual risk taking among HIV
-infected and uninfected homosexual men. AIDS
7. Fisher J, Smith L. Secondary prevention of HIV
infection: the current state of prevention for positives. Curr Opin HIV AIDS
8. Salomon J, Hogan DR, Stover J, Stanecki KA, Walker N, Ghys PD, Schwartländer B. Integrating HIV
prevention and treatment: from slogans to impact. PLoS Med 2005; 2:e16.
9. Granich R, Gilks CF, Dye C, De Cock KM, Williams BG. Universal voluntary HIV
testing with immediate antiretroviral therapy
as a strategy for elimination of HIV transmission
: a mathematical model. Lancet 2008; 373:48–57.
10. Dodd P, Garnett GP, Hallett TB. Examining the promise of HIV
elimination by ‘test and treat’ in hyperendemic settings. AIDS
11. United Nations. 2008 update on the global AIDS epidemic.
12. UNAIDS. South Africa – progress towards universal access.
13. Pettifor A, Kleinschmidt I, Levin J, Rees HV, MacPhail C, Madikizela-Hlongwa L, et al
. A community-based study to examine the effect of a youth HIV
prevention intervention on young people aged 15-24 in South Africa
: results of the baseline survey. Trop Med Int Health 2005; 10:971–980.
14. Eaton L, Flisher AJ, Aarø LE. Unsafe sexual behaviour in South African youth. Soc Sci Med 2003; 56:149–165.
15. Pettifor A, Rees HV, Kleinschmidt I, Steffenson AE, MacPhail C, Hlongwa-Madikizela L, et al
. Young people's sexual health in South Africa
prevalence and sexual behaviors from a nationally representative household survey. AIDS
16. Human Sciences Research Council. South African National HIV Prevalence, Incidence, Behavior, and Communication Survey, 2008: a turning tide among teenagers?
Cape Town, South Africa
: HSRC Press; 2008.
17. Hargreaves J, Bonell CP, Morison LA, Kim JC, Phetla G, Porter JD, et al
. Explaining continued high HIV
prevalence in South Africa
: socioeconomic factors, HIV
incidence and sexual behaviour change among a rural cohort, 2001-2004. AIDS
18. Janssen RS, Holtgrave DR, Valdiserri RO, Shepherd M, Gayle HD, De Cock KM. The serostatus approach to fighting the HIV
epidemic: prevention strategies for infected individuals. Am J Public Health 2001; 19:1019–1024.
19. Kennedy C, O'Reilly K, Medley A, Sweat M. The impact of HIV
treatment on risk behaviour in developing countries: a systematic review. AIDS
Care 2007; 19:707–720.
20. Sarna A, Luchters SM, Geibel S, Kaai S, Munyao P, Shikely KS, et al
. Sexual risk behaviour and HAART
: a comparative study of HIV
-infected persons on HAART
and on preventive therapy in Kenya. Int J STD AIDS
21. Moatti J, Prudhomme J, Traore DC, Juillet-Amari A, Akribi HA, Msellati P, Côte d'Ivoire HIV
Drug Access Initiative Socio-Behavioural Evaluation Group. Access to antiretroviral treatment and sexual behaviours of HIV
-infected patients aware of their serostatus in Côte d'Ivoire. AIDS
22. Bateganya M, Colfax G, Shafer LA, Kityo C, Mugyenyi P, Serwadda D, et al
. Antiretroviral therapy
and sexual behavior
: a comparative study between antiretroviral- naive and -experienced patients at an urban HIV
care and research center in Kampala, Uganda. AIDS
Patient Care STDs 2005; 19:760–768.
23. Bunnell R, Ekwaru JP, Solberg P, Wamai N, Bikaako-Kajura W, Were W, et al
. Changes in sexual behavior
and risk of HIV transmission
after antiretroviral therapy
and prevention interventions in rural Uganda. AIDS
24. McClelland R, Graham SM, Richardson BA, Peshu N, Masese LN, Wanje GH, et al
. Treatment with antiretroviral therapy is not associated with increased sexual risk behavior in Kenyan female sex workers
25. Marcellin F, Bonono CR, Blache J, Carrieri MP, Spire B, Koulla-Shiro S, the EVAL Study Group. Higher risk of unsafe sex and impaired quality of life among patients not receiving antiretroviral therapy
in Cameroon: results from the EVAL survey (ANRS 12-116). AIDS
26. Dia A, Marcellin F, Bonono R-C, Boyer S, Bouhnik A-D, Protopopescu C, et al
, the EVAL Study Group. Prevalence of unsafe sex with one's steady partner either HIV
-negative or of unknown HIV
status and associated determinants in Cameroon (EVAL ANRS12-16 survey). Sex Transm Infect 2010; 86:148–154.
27. Golub J, Pronyk P, Mohapi L, Thsabangu N, Moshabela M, Struthers H, et al
. Isoniazid preventive therapy, HAART
and tuberculosis risk in HIV
-infected adults in South Africa
: a prospective cohort. AIDS
28. Omar TSS, Tshabangu N, Golub J, Hanrahan C, Gray G, Struthers H, et al
. Progression and regression of premalignant cervical lesions in HIV-infected women from Soweto: a prospective cohort 17th Conference on Retroviruses and Opportunistic Infections
. San Francisco; 2010.
29. South African Clinicians Society guidelines: antiretroviral therapy
in adults. S Afr J HIV Med
30. World Health Organization. Towards universal access: scaling up priority HIV/AIDS interventions in the health sector; 2008
31. Lurie M, Pronyk P, de Moor E, Heyer A, de Bruyn G, Struthers H, et al
. Sexual behavior
and reproductive health among HIV
-infected patients in urban and rural South Africa
. J Acquir Immune Defic Syndr 2008; 47:484–493.
32. Allison P. Logistic regression using the SAS system: theory and application
. Cary, North Carolina, USA: SAS Institute; 1999.
33. Allison P. Fixed effects regression methods for longitudinal data using SAS
. Cary, North Carolina, USA: SAS Institute Inc.; 2005.
34. Luchters S, Sarna A, Geibel S, Chersich MF, Munyao P, Kaai S, et al
. Safer sexual behaviors after 12 months of antiretroviral treatment in Mombasa, Kenya: a prospective cohort. AIDS
Patient Care STDs 2008; 22:587–594.
35. Eisele T, Mathews C, Chopra M, Lurie MN, Brown L, Dewing S, Kendall C. Changes in risk behavior among HIV
-positive patients during their first year of antiretroviral therapy
in Cape Town South Africa
Behav 2008; 13:1097–1105.
36. Wanyama J, Kasibante D, Kayiira P, Wandera B, Kambugu A. Sexual behavior of HIV+ persons before and after 1 year of ART at the infectious diseases clinic, Kampala, Uganda
. 17th Conference on Retroviruses and Opportunistic Infections
. San Francisco; 2010.
37. Kaida A, Gray G, Bastos FI, Andia I, Maier M, McIntyre J, et al
. The relationship between HAART
use and sexual activity among HIV
-positive women of reproductive age in Brazil, South Africa
, and Uganda. AIDS
Care 2008; 20:21–25.
38. Diabaté S, Alary M, Koffi CK. Short-term increase in unsafe sexual behaviour after initiation of HAART
in Côte d'Ivoire. AIDS
39. Lurie M, Williams BG, Zuma K, Mkaya-Mwamburi D, Garnett GP, Sweat MD, et al
. Who infects whom? HIV
-1 concordance and discordance among migrant and nonmigrant couples in South Africa
40. Eisele T, Mathews C, Chopra M, Brown L, Silvestre E, Daries V, Kendall C. High levels of risk behavior among people living with HIV
initiating and waiting to start antiretroviral therapy
in Cape Town South Africa
Behav 2008; 12:570–577.
41. Reynolds S, Makumbi F, Kagaayi J, Nakigozi G, Galiwongo R, Quinn T, et al
. ART reduced the rate of sexual transmission of HIV among HIV-discordant couples in rural Rakai, Uganda. 16th Conference on Retroviruses and Opportunistic Infections.
42. Carpenter L, Kamali A, Ruberantwari A, Malamba SS, Whitworth JA. Rates of HIV
-1 transmission within marriage in rural Uganda in relation to the HIV
sero-status of the partners. AIDS
43. Allen S, Meinzen-Derr J, Kautzman M, Zulu I, Trask S, Fideli U, et al
. Sexual behavior
discordant couples after HIV
counseling and testing. AIDS
44. Protopopescu C, Raffi F, Roux P, Reynes J, Dellamonica P, Spire B, et al
, ANRS CO8 (APROCO-COPILOTE) Study Group. Factors associated with nonadherence to long-term highly active antiretroviral therapy
: a 10 year follow-up analysis with correction for the bias induced by missing data. J Antimicrob Chemother 2009; 64:599–606.
45. Mayer K, Venkatesh KK. Antiretroviral therapy for HIV prevention: current status and future prospects
. Am J Public Health
2010 (in press).
46. Burns D, Dieffenbach CW, Vermund SH. Rethinking HIV-1 prevention
. Clin Infect Dis
2010 (in press).