The impact of antiretroviral treatment on fertility intentions in South Africa

Polly Clayden, HIV i-Base

There is limited information about the impact of expanding access to HAART in settings with limited resources and large epidemics on women’s reproductive decisions and outcomes.

Angela Kaida showed findings from an investigation conducted to assess whether the use and duration of HAART was associated with: fertility intentions, contraception use and method preference, and the incidence of live birth, among women attending the Perinatal Research Unit (PHRU) in Soweto, South Africa.

The study was cross-sectional and used an interviewer-administered survey and a case note review. A total of 751 women, aged 18-49, took part. Of these, 253 had received HAART for a median duration of 31 months. The mean CD4 count was 406 cells/mm3 and 81% had undetectable viral load (group 1). A further 249 women were also HIV-positive but HAART-naive, with a mean CD4 count of 351 cells/mm3 (group 2). A reference group included 249 HIV-negative women (group 3).

Multivariate analysis (n=674) revealed HIV-positive women were nearly 60% less likely to report fertility intentions than HIV-negative women but the difference between those receiving treatment and naive women was modest. With HIV-negative women as reference, the investigators reported adjusted odds ratio (AOR) 0.35 (95%CI 0.21-0.60) and AOR 0.4 (95% CI 0.23-0.69) for women HAART-naive and receiving HAART respectively.

When the investigators looked at the prevalence of contraceptive use among non-pregnant sexually active women (n=563) in this cohort, they found that use was high–nearly 80%, compared to an average of just over 60% among South African women in general. Women receiving HAART were significantly more likely to use contraception: 86% of women receiving HAART, 82% of HAART-naive women and 69% of HIV-negative women reported contraceptive use, p<0.001. Multivariate analysis, compared to HIV-negative women, found AOR 1.59 (95% CI 0.88-2.85) and AOR 2.40 (95% CI 1.25-4.62) for women HAART-naive and receiving HAART respectively. The investigators also noted that women receiving HAART were more likely to use dual contraception.

Finally Dr Kaida presented preliminary data from an assessment of lifetime incidence of live birth by time period. For this analysis each participant (n=748) contributed woman-years of follow up based on dates of HIV diagnosis and starting HAART (for those who had). With the HIV-negative time period as a reference, this analysis showed a 69% higher incidence of live birth in the HAART naive time period than the HIV-negative period—adjusted relative risk (ARR) 1.69 (95%CI 1.48–1.93)—but 66% lower in the period when women received HAART, ARR 0.34 (95% CI 0.23–0.49).

The investigators suggested that this study highlights the potential value of improved integration between HIV prevention, testing and HAART services with sexual and reproductive health programming.


Kaida A et al. The impact of expanding access to HAART on fertility intentions, contraceptive use and fertility among women in an HIV hyper-epidemic setting. 1st International Workshop on HIV and Women. 10–11 January 2011, Washington. Oral abstract O_09.

Links to other websites are current at date of posting but not maintained.