Microfinance project reduces HIV risk in South African women, gold standard trial shows

A microfinance and training project designed to empower South African women was associated with a significant reduction in HIV risk behaviour and partner violence over two years of follow-up, delegates heard at the XVII International AIDS Conference in Mexico City on Monday.

The study, a randomised cluster trial in South Africa, adds to the currently limited body of evidence concerning the effects of gender and economic-related interventions on HIV vulnerability and risk behaviour.

National HIV programmes and governments have been urged by the UN Secretary-General to intensify the focus on gender inequities and sexual norms, and their roles in increasing HIV risk and vulnerability.

But the dilemma facing those who want to scale up such interventions is the limited evidence of what works: there have been few randomised studies and a systematic review led by USAID could identify only eight intervention studies with a gender focus that measured health outcomes such as HIV infections or HIV testing and which had been systematically evaluated, either in a trial or other rigorous methodology.

Microfinance schemes aimed at the poorest members of society have spread quite widely in the developing world since the system of very small loans without security was first pioneered in the developing world by the Grameen Bank in Bangladesh in the 1970s.

Microfinance has been suggested as a means of empowering economically disadvantaged women in order to reduce their risk of HIV infection, since studies have shown that women are often vulnerable to HIV infection due to financial dependence on men and a subsequent inability to negotiate safer sex.

The IMAGE microfinance study was conducted in eight rural communities in the Limpopo province of South Africa as a cluster randomised trial, in which four communities immediately received the intervention and four received the intervention at a later stage, in order to observe differences between the communities.

The study recruited the poorest women in each community and offered small loans to start businesses such as buying and selling produce, selling used or new clothes, providing child care services or running food stalls. A total of 1750 loans were granted, totalling $290,000. The repayment rate was 99.7%.

The study intervention consisted of ten compulsory one-hour training sessions integrated into the fortnightly loan centre meetings. The training sessions addressed topics such as HIV, sexual violence, gender inequality and communication skills. Eight hundred and sixty participants were compared with a control group of women who did not receive the intervention.

The second part of the intervention consists of a community mobilisation process in which leaders are selected from the loan groups and given leadership training, and then develop local action plans based on the concerns of local women.

The findings reported at the International AIDS Conference only assess the effects of the first phase of the intervention among direct participants. Subsequent research will look at the effects of the second phase of the intervention on sexual behaviour and women’s empowerment in the wider participating community, focusing especially on sexual behaviour in young women aged 14-35.

The results of the first phase were assessed according to a number of parameters: economic well-being, empowerment, intimate partner violence and HIV-related variables.

Two years after the beginning of the intervention the study showed a range of positive outcomes among women who took part:

• Economic well-being: A significant increase in household asset value was noted (adjusted relative ratio 1.15, 95% confidence interval 1.04-1.78), together with a trend towards a higher household savings rate (ARR 1.84, 95% CI 0.77 – 4.37). No differences in school enrolment and food security were noted.
• Empowerment: the only significant variable was communication with household members. Women who participated in the microfinance/training scheme did not show significantly greater self-confidence or better communication with a partner when compared with the control group.
• Intimate partner violence: A statistically significant 55% reduction in intimate partner violence was noted over the two-year follow-up period.
• HIV-related behaviours: Unprotected sex with non-spousal partner at last intercourse declined significantly (by 24%) in the intervention group (ARR 0.76, 95% CI 0.60-0.96). Participants were also significantly more likely to access voluntary counselling and testing, and reported higher levels of HIV-related communication with their sexual partners. However an incidence study in participating communities showed no overall decline in HIV incidence during the study period.

The study intervention is now being scaled up to a wider range of communities in Limpopo province, with the result that the unit cost per client of the programme has fallen from $43 per client to $13 per client when 4,500 households are reached. The investigators intend to reach 15,000 households in the period 2008-2010.

“To have a 50% reduction in gender-based violence is absolutely tremendous,” said Karen Hardee of Population Action International, who presented an overview of gender-focused interventions during the same conference session.

But, she warned, “changing gender norms is time-intensive,” and changes in gender norms may take time to translate into behavioural changes. She warned that large studies such as IMAGE and Stepping Stones may still be too short to detect changes in HIV incidence either among participants or at the community level.