South Africa: women may have higher rates of HIV than men because fewer men are on treatment

South African researchers presenting a study at the recent Conference on Retroviruses and Opportunistic Infections (CROI 2018) have suggested that the reason the extremely high rates of HIV infection in women are stubbornly refusing to decline is, seemingly unfairly, because women are more likely to be diagnosed and take antiretroviral therapy than men.

While this may save the lives of the ones who do have HIV, in terms of those who don’t, the higher rates of viral suppression in women largely benefit men. The uptake of circumcision may also be contributing to an observed fall in HIV incidence in men.

Alain Vandormael of the University of KwaZulu-Natal said that getting more, and younger, men on to HIV treatment was essential in order to reduce the persistently high HIV incidence in young women, especially if combined with prevention measures that offer women the same degree of protection as medical male circumcision does to men.

The AHRI cohort – latest data

Vandormael was presenting the most recent findings from a long-lasting cohort study in north-eastern KwaZulu-Natal conducted by the African Health Research Initiative (AHRI). It was this cohort that in 2012 provided the first evidence from a low-income setting that getting a sufficiently high proportion of the population virally suppressed was starting to bring down the HIV infection rate.


detectable viral load

When viral load is detectable, this indicates that HIV is replicating in the body. If the person is taking HIV treatment but their viral load is detectable, the treatment is not working properly. There may still be a risk of HIV transmission to sexual partners.

voluntary male medical circumcision (VMMC)

The surgical removal of the foreskin of the penis (the retractable fold of tissue that covers the head of the penis) to reduce the risk of HIV infection in men.


The surgical removal of the foreskin of the penis (the retractable fold of tissue that covers the head of the penis) to reduce the risk of HIV infection in men.

person years

In a study “100 person years of follow-up” could mean that information was collected on 100 people for one year, or on 50 people for two years each, or on ten people over ten years. In practice, each person’s duration of follow-up is likely to be different.

statistical significance

Statistical tests are used to judge whether the results of a study could be due to chance and would not be confirmed if the study was repeated. If result is probably not due to chance, the results are ‘statistically significant’. 

Because this area has comprehensive HIV surveillance, part of it was also chosen to host ANRS 12249, the world’s first study of the population-level effects of “Test and Treat” as a strategy.

In the AHRI cohort, people are proactively tested by community health workers, who visit each household in the catchment area, using dried blood spot technology, once a year. Between 2004 and 2015, 38,597 people have been tested for HIV at least once and exactly 17,400 testers (45%), whose first test was HIV negative and who have taken at least one more HIV test since then, form the “incidence cohort”. Of these, 9908 (57%) are women.

Among men, there have been 641 new diagnoses of HIV in 28,006 person-years (overall annual incidence, 2.3%) and in women, 1915 in 40,182 person-years (4.8%).

HIV incidence was not only lower in men than in women generally, it also declined from 2.4% in 2005 to 2.05% in 2010 and then to 0.95% in 2015. From 2010 onwards this fall was statistically significant.

In contrast HIV incidence in women during the same period increased, from 4.2% it 2005 to 4.65% in 2010 and then 5.1% in 2015. This nearly reaches statistical significance. There are signs of a levelling-off in the incidence rate from 2013 to 2015, but not a decline.

Why, among a population that has seen antiretroviral therapy (ART) coverage of those in care rise by over 60% in the last five years, have men seen their rate of HIV infection fall by over half, while among women there has been an increase?

One explanation has been that age differences are partly to blame. It is certainly true that HIV annual HIV incidence is still disproportionately high among young women: currently 5.4% of girls aged 15-20 are infected with HIV per year, 7% aged 20-25 and 6% aged 30-35. By the time women are 30, a huge proportion of them have HIV – more than half in some areas.

Annual HIV incidence in men, on the other hand, is very low in youth – just 0.7% in boys aged 15-20 – and does not start to exceed that in women until the ages of 35-40, when it is 3% in men and 2.2% in women.

More women on treatment means fewer men infected

However, the researchers did not believe age-mixing was entirely the reason for the much higher rates in young women. They also thought it might, in recent years, have to do with the fact that more women than men are on treatment. Women are tested for HIV earlier – often in an antenatal setting – and are put on ART earlier.

Between 2010 and 2015 the proportion of women with HIV who were on treatment – including those not in regular care - increased by 66% from 29% to 49%. Among men, however, it only increased from 26% to 38% – a 46% increase.

Did the increase in the number of people on ART have a population-level effect on the overall infectiousness of the HIV-positive population? Among women, yes it did. The best way to evaluate this is to estimate the proportion of the entire population who have a detectable viral load, not just those who are known to have HIV.

In women, the proportion of the entire population with a detectable viral load did decline between 2011 and 2014, from 21.5% to 19.5%. This might not look a lot in absolute terms but does represent a 9.3% decline. In men, on the other hand the proportion with a detectable viral load actually increased during these three years from 14% to 17% – a 21% increase.

In other words, while age mixing is part of the explanation, another part is simply that young women are much more likely to be diagnosed and on HIV treatment than young men – and although in absolute terms fewer men have HIV, a higher proportion of them are infectious.

During this period the proportion of men who were circumcised also grew and though Alain Vandormael did not give figures for this, in a study from KwaZulu-Natal last year, while 51% of youth aged 15-19 were circumcised, less than 20% of men aged over 35 were. This may partly explain the big differential between female and male HIV infection rates in younger people.

Vandormael said that to reduce the still exceptional rates of HIV infection in young women, two things needed to happen. Firstly, men, and especially young men, not only needed to be tested regularly for HIV but also better linked to care and treatment. And secondly, women badly needed an equivalent to medical male circumcision for HIV prevention. The vaginal ring, while not lifelong in its effect, might represent such an opportunity, Vandormael said.


Vandormael A et al. Sharp decline in male HIV incidence in a rural South African population (2004–2015). Conference on Retroviruses and Opportunistic Infections, Boston, abstract 46, 2018.

View the abstract on the conference website.

Watch the webcast of this presentation on the conference website.