We now know the answer after more than 25 years of guessing, but will women believe it?
The world has spent nearly a quarter of a century wondering whether Africa's most widely used birth control method could make women more likely to contract HIV. Now, new research, conducted in four countries, including South Africa, has solved the riddle.
The three-month shot Depo-Provera does not increase a woman's risk of HIV infection, prove the results of the Evidence for Contraceptive Options and HIV Outcomes (Echo) study. The findings were released at the South African Aids Conference in Durban Thursday and published in The Lancet.
For decades, it was there like a spectre among the data - something researchers thought they saw but couldn't entirely be sure.
And it wasn't just in the lab.
From South Africa to Tanzania, studies kept picking up what seemed like an association between the widely used three-month birth control shot marketed as Depo-Provera and HIV infection.
A 2016 review of nearly three dozen studies published in the journal Aids suggested Depo-Provera users could be 20% to 60% more likely to contract the virus compared to women who didn't rely on this hormone-based contraceptive method.
The problem was, none of this research proved it. Why? Because it was never designed to do that.
Scientists couldn't tell if what they thought they saw in the numbers was right - whether it was the shot or if they were picking up on something else that was putting women at risk such as the lower rates of condom use that a 2005 study by the Guttmacher Institute detected among some Depo users.
The findings of the Echo study are the product of a three-year study conducted among about 7 800 women in eSwatini, South Africa, Kenya and Zambia.
As part of this, women wanting birth control agreed to be randomly assigned one of three contraceptive methods - a three-monthly shot of Depo-Provera, a copper implant inserted into the uterus; or a small, hormone-based upper arm implant known as Jadelle. Although women were allowed to change the type of contraception they were assigned to if they wished, most didn't. And because scientists believed it was unethical to withhold contraception from women who wanted it, everyone got one of the three choices.
Randomisation like this helps ensure that any factors, such as age, that could affect the study's results are spread equally among all groups, making these kinds of "gold standard" studies the best at determining cause and effect relationships.
The Echo found no substantial difference in the rates of HIV infection among the three groups.
So what should South African women know?
"From what we can see there is no increased risk from any of those three methods in terms of increasing your chances of contracting HIV", says executive director of the Wits Reproductive Health Institute Helen Rees. Rees is also a member of the five-person committee that led the study.
"However, none of these methods will protect you against HIV... and you still need to protect yourself from HIV infection", she explains. "But if you want to protect yourself against unplanned pregnancy, these methods are a real option for you."
It's simple messages like this that will now need to filter down to women and communities in languages and ways they can understand.
But their voices will also have to be represented in the high-level discussions that will follow Thursday's results.
Historically, the world has done neither well, and there may be a price to pay for that.
In October 2011, The Lancet published a study that suggested a possible link between Depo-Provera and HIV risk. It prompted a high-level World Health Organisation (WHO) committee to see if it should issue warnings to doctors and women about the contraception that forms the backbone of birth control programmes in much of the Global South.
But with the kind of conclusive evidence Rees' trial provided this week, the WHO was left without much to say and not a very clear way of saying it: It advised women using Depo-Provera to also use condoms.
Activists were not impressed.
In reality, most women who opt for long-acting contraception like Depo-Provera do so because they are often unable to negotiate condom use, co-director of the international organisation Aids-Free World Paula Donovan pointed out at the time.
Lillian Mworeko is the East Africa regional coordinator of the International Community of Women Living with HIV/Aids.
At the time, Mworeko was one of the few women with HIV who was invited to the high-level WHO consultations on Depo-Provera and HIV. At one point, the Ugandan activist and others were made to sign a confidentiality agreement that would have prevented them from reporting back to their networks back home, Donovan argued.
Read more: "I'm an activist and woman living with HIV. This is why the study's results scare me"
Mworeko called the WHO out for issuing a lack of clear guidance at the time: "Women at high risk of or living with HIV still have a right to informed consent, which includes the right to information that affects their health."
By 2017, the WHO had changed its tune, telling women that for most, Depo-Provera works well enough, but for some, it may be risky.
It would be another four years before Rees could fundraise the US$50-million (R746-million) needed to finance Echo, the world's first study that investigated whether Depo-Provera fueled HIV infections. In total, Rees and her fellow researchers spent 15 years planning - and finding money - for the Echo study.
"Was it easy to raise money for this trial?" she asks. "The answer is no, it wasn't, and we have to ask why that is."
Rees told Bhekisisa in November 2018: "'If there was a product that men were using that had a question like this hanging over it. We would have had an answer by now."
We have the science, what now?
By 2017, the WHO had begun to tell women that Depo-Provera could make them more likely to contract HIV in hopes that it would allow those already at high risk for HIV - including many women in South Africa - make an informed decision about their choice. In practice, however, few women in the Global South would have been given a panoply of other options.
The WHO is preparing to issue updated guidance on Depo-Provera by late August, WHO spokesperson Tarik Jašarević told Bhekisisa. As part of this, a 28-person expert panel, including clinical experts and outreach workers from countries such as South Africa and Zimbabwe will review the evidence. Their nominations were put forth for public comment in May.
But only one of the participants is a woman openly living with the virus, according to biographies posted online.
Meanwhile, years without clear evidence and waffling about what to say to women about contraception and HIV risk, and how to say it, may have taken its toll.
Echo's findings come amid ongoing contraception stock outs that have run into their second year: "At the moment with stock outs, I could be given Depro-Provera one month if it's available and the next time I come to the clinic I might only be able to get [birth control pills]. People are being switched between methods without even being consulted on the side effects", activist Nomfundo Eland says.
Weeks ahead of this week's release of the Echo results, The Lancet accidentally released the Echo research online briefly. Both the journal's media relations manager Emily Head and Rees confirmed to Bhekisisa that the early release was an administrative error.
Rees explains: "It was absolutely nothing more sinister than that."
But longtime HIV activist Nomfundo Eland argues it's raised suspicions among women who may already be sceptical of the findings despite the study's rigorous methods.
"We are going to have a hectic time [getting] women to believe it, especially women and activists who have been following the Echo agenda because the results were leaked", she says. "We are asking ourselves questions - who leaked it? For what purposes? What information got out, and who has it?"
Eland warns activists also remain concerned that the early release led some study sites to rush feedback of results to participants, leaving them confused. Rees, however, says that although the release did push some time frames forward, each of the 12 sites had already developed its own plan to let women know what the study found at the same time results were made public.
In some cases, sites let women know of the findings before this week.
But Rees admits some scepticism is perhaps to be expected.
"We hope that this study for most people will be massively reassuring. This was the gold standard in terms of studies, and it was a study that was very well executed", she explains.
"But this is an issue that has been discussed backwards and forwards for years because no one was able to do a definitive study until now. That's left ambiguity, and when you have ambiguity, people build up a view."
Rees concludes: "There was certainly many people of the view, looking at the data that existed [before]... that there was a trend towards increased risk.," Rees continues.
That some people won't believe the results even three years of research may be unavoidable.
It's probably inevitable because of the strength of view that has been built up over the years."
Echo solved one riddle, but it left many more.
About one in 20 women contracted HIV during over the course of the study's three years - not because of their birth control, but because they face the same social and economic issues that lead 1300 young women in South Africa to contract the virus every week, shows the country's latest household survey.
We need better and more contraception choices for women, Rees and Eland warn.
And more ways to reach women with the tools they need to prevent HIV.
"If women go for family planning, they get family planning", Rees explains. "If they go into HIV services, they get HIV services.
"We're not integrating our services... these are women who are presenting themselves and who are sexually active - they need both. Together."