In the 30 plus years since HIV was first described, there’s been an omnipresent desire to identify the “bad actors” that are responsible for HIV becoming an epidemic. This is true nowhere more than in sub-Saharan Africa (SSA). Much blame over the years has been laid at the feet of those who are perceived as morally inferior rather than dealing with the reality that the reason HIV in SSA has become more prevalent than in other parts of the globe are complex and not fully understood.
In the scientific and planning literature, the language of moral judgment has shifted over time towards calling out groups as “drivers of the epidemic” on whom the blame can be heaped. In large part, these “drivers” are nothing of the sort and are merely boogeymen that can conveniently be utilized to justify a sexual morality that has little to do with public health. As our knowledge of HIV has expanded and as more scientific research has been conducted, it’s unfortunate that many of these same “drivers” remain core parts of the prevention agenda and continue to divert prevention resources away from interventions that have a proven evidence base and effectiveness outcomes.
This post is meant to illuminate a few of the false narratives that have been constructed to explain the HIV epidemic in moralistic terms but where observational evidence has shown that the populations being identified are no more the drivers of the epidemic than the general population. I do not mean to suggest that targeted prevention services are unnecessary nor that all “drivers” in the scientific literature are equally deficient of evidence – but that the description of groups as main culprits of HIV is often time simple re-branding of moralistic attitudes.
“Sugar Daddies” (Aka Intergenerational Sex)
Looking at HIV strategy documents in Namibia, South Africa, and other places will show that “Age Discordant Relationships” or “Intergenerational Sex” are drivers of HIV incidence in SSA. These arguments have been made largely based on the evidence that teen girls have higher HIV incidence and prevalence rates than teen boys in the same area. While it’s clear that young girls are at much higher risk of acquiring HIV, a recent study published at the Conference on Retroviruses and Opportunistic Infections (CROI) is the first that actually studies the relationship between age-discordant relationships and HIV incidence.
The study – Age-Disparate Relationships and HIV Incidence Amongst Rural South African Women – utilized data from the Africa Centre for Health and Population Studies database in a high-prevalence rural district of South Africa. By looking at all sexually-active women aged 15-29 and 30-57 who received annual HIV tests between 2005 and 2012 and reported on the age disparity of their most recent relationships at baseline the study was able to identify incident HIV infections and correlate them to the age-disparity of their relationships.
The results for both groups are shown below:
The data is relatively clear that the vast majority of transmissions to women 15-29 occurred in relationships where the age differences were relatively small. Although not shown here, this held true even when smaller 5-year age groups were analyzed and when modelling whether relationship types were changing over time.
The authors of the study conclude that age-disparities in relationships are present, but are not actually risky for increased HIV transmission. Importantly, amongst women 30 and older, the data showed that larger age disparities were actually more protective than smaller age-disparities indicating that stigmatizing age-discordant relationships could actually be harmful to women if it prevents them from engaging in them.
Now, this is only one study in one rural district in South Africa and there are good reasons to seek to confirm the findings with data sets in urban environments, wealthier environments, and other countries. However, it is odd that much planning time and expenditure of valuable prevention resources are being consumed by prevention programs targeted at relationships for which there is little to no supporting data outside of theoretical models. There may be reasons to discourage “sugar daddy” relationships based on gender equality and power dynamics in those relationships (assuming there is an evidence base for this). However, that those relationships may or may not be desirable does not justify – without evidence – identifying them as risky in terms of HIV transmission.
The Non-Monogamists (Aka Multiple Concurrent Partnerships)
Multiple concurrent partnerships (MCPs) has become a favored go-to for arguing why HIV prevalence in SSA is so much higher than in other areas of the globe (see the above Namibia and South Africa plans as well as here and here). The logical argument in favor of MCPs as drivers is that they create tighter networks of individuals through which HIV can spread rapidly when there’s a new incident infection within the network. This correlates with evidence that HIV is most infectious in the period shortly after initial seroconversion.
Unfortunately for the proponents of the MCP hypothesis, the reality on the ground doesn’t match up with the models used to argue that MCPs are major drivers for the epidemic. The original model by Morris and Kretzschmar that brought the MCP hypothesis to the fore was published in 1997. It made unrealistic assumptions incompatible with realities on the ground to achieve its results – mostly relating to the frequency of sex, the assumed transmission rates, and the levels of concurrency in African countries. An excellent critique of the model was produced by Sawers and Stillwaggon. Essentially, the model assumed that each person in a concurrent partnership had sex with each partner every single day (thus a person with four partners was having sex four times per day every single day throughout each partnership), that the transmission rate for HIV was at least 50 times higher than best estimates, and by assuming that 50% of sexual relationships are concurrent – rates that are far above any data on concurrency anywhere in the world.
Over time, the model has been adapted and updated by Morris and Kretzschmar as well as by others (see Eaton, et al.). Yet severely flawed assumptions remain that remove the models from the realm of reality and place them into the realm of pre-determined bias. Proponents of the MCP hypothesis have argued that the models are merely “proof of concept” models showing that simulated HIV epidemics can theoretically spread faster and grow larger in populations with high MCPs than in populations where serial monogamy is the norm. This is – of course – all well and good except that the true test of any model’s validity is whether it can align with reality once real world data are used in place of made up assumptions. On these grounds, the models continue to fail miserably. Sawers 2013 review of the evidence and modeling of concurrency is essential reading on the topic.
The reality is that human sexual behaviors are not significantly different between SSA and the rest of the world and cannot explain the differences in scales of the HIV epidemics. As discussed in the Sawers 2013 article, based on the UNAIDS consensus definition on concurrency, rates of concurrency in SSA are not significantly different than the United States. If concurrency is one of the main driving factors for the scale of the epidemic in SSA, we should be able to see it in the data and not just in severely flawed mathematical models designed specifically to show that concurrency spreads HIV faster than a non-existent world where people only engage in serial monogamy. We can’t see it and it is well past time that this line of argument be put to rest unless and until new real world data are gathered to show how MCPs in SSA differ so drastically from MCPs in the US, Europe, Asia, or elsewhere that can explain exponentially higher prevalence rates.
The “Bad” Sexually Active HIV Positive Individuals
It’s long been hypothesized that there are a significant number of HIV positive individuals who are willfully or recklessly spreading HIV to others. This fear is largely the basis of HIV criminalization laws that have become common throughout the world. These laws – put in place supposedly to punish the rash of “bad men” spreading HIV – have ironically been utilized more often to prosecute women in SSA who are far more likely to know their HIV status. Yet the reality of these “bad men” is that they have only ever been anecdotes. In effect, they are not a public health threat given the context of the total epidemic and focus on the vanishingly few individuals detracts from overall prevention goals and increases stigmatization of the disease.
In reality, people who are aware of their HIV positive status are far more likely to reduce risky sexual behaviors than people who are unaware of their status. In addition, we now know that HIV positive individuals who are on effective treatment with undetectable viral loads are essentially non-infectious. The HPTN 052 study – which initially established the evidence that there was a 96% reduction in HIV transmissions in sero-discordant heterosexual couples – has now been joined by the PARTNER study which found that HIV positive individuals with an HIV viral load below 200 do not transmit HIV during vaginal or anal intercourse.
The reality remains what it has always been. The vast majority of HIV transmissions take place between individuals who are unaware of their HIV status. For further evidence, see here.
HIV prevention programs have been paralyzed for years. Intractable stigma – which myths such as those above serve to enhance – and insufficient resources continue to undermine HIV prevention programs. Diverting prevention resources into programs that actively enhance stigma – without evidence – is damaging to the programs.
While it remains unknown at present why HIV in SSA has become much more endemic than elsewhere, differences in human sexuality do not explain it. A poster presentation at CROI – Transmission Rates and Not Sexual Contact Patterns Drive HIV Epidemic Intensity in Africa – shows that it’s far more likely that HIV transmission rates are more likely to blame but for unknown reasons – potentially relating to co-infections or genetics differences.
In any event, if the proponents of models for HIV that rely on substantial differences in human sexuality wish to continue to consume critical prevention resources in SSA, it’s well past time that they put forward compelling scientific evidence that their hypothesis is anything more than simple moral judgment and dislike. As Francois Venter recently said in a discussion of the sugar daddies data:
Well, then lets call it what it is – a moral judgment, not a public health intervention. I’d prefer we kept them apart. I think too much of HIV prevention is based on people’s discomfort with sex as a human need, rather than evidence. If people don’t like sugar daddies, concurrency, pornography, unmarried people, sex work, so be it, but stop hiding behind HIV prevention aprons.
The views reflected in this expert column are those of the individual authors and do not necessarily represent those of the O’Neill Institute for National and Global Health Law or Georgetown University. This blog is solely informational in nature, and not intended as a substitute for competent legal advice from a licensed and retained attorney in your state or country.