Monday January 12 2015
The Depo-Provera injection lasts for 12 weeks
"Contraceptive injections moderately increase a woman's risk of becoming infected with HIV," The Guardian reports.
The headline was prompted by an analysis of 12 studies that looked at whether the use of hormonal contraception, such as the oral contraceptive pill, increases the risk of contracting HIV.
All of the studies involved were conducted in sub-Saharan Africa in low- and middle-income countries.
Researchers found a link between a common injectable form of contraception called depot medroxyprogesterone acetate (Depo-Provera) and the risk of HIV. No link was found with other types of hormonal contraception.
But these results do not prove the depot injection directly increases the risk of HIV. The studies included varied in their design and methods, and have several potential sources of bias.
Any link could be down to behavioural patterns rather than medical reasons. For example, women who know they have an effective long-term contraceptive may forget about the risks of sexually transmitted infections.
Hormonal contraception, including injections or oral tablets, can be an extremely effective form of contraception. But it won't protect you against sexually transmitted infections.
It is worth discussing with your health professional and making sure you are using the method that is most effective, convenient and safest for you, depending on your circumstances.
Where did the story come from?
The study was carried out by researchers from the University of California and received no financial support.
It was published in the peer-reviewed medical journal, The Lancet.
The Mail Online correctly reports the main findings of this study, but would benefit from highlighting that the findings do not prove a causal association between the depot injection and HIV risk, a point clearly made by the researchers in the original publication.
The Guardian's reporting of the study is more measured and highlights how for women in poorer countries, an unwanted pregnancy may pose a greater threat to health and wellbeing than HIV. Rates of maternal death occurring during or shortly after pregnancy remain high in many sub-Saharan countries.
What kind of research was this?
This was a systematic review that aimed to search the global literature to find studies examining whether the use of hormonal contraception, such as the oral contraceptive pill or contraceptive injections, increase the risk of contracting HIV.
The researchers say previous study into whether there could be an associated risk has been inconsistent. They pooled the results of different studies in a meta-analysis.
A systematic review and meta-analysis is the best way of identifying and looking at all evidence that has addressed the particular question of interest.
But this type of research is always going to have some limitations reflecting the strength and quality of the underlying studies being reviewed.
It is unlikely a trial would be conducted that would allocate women to hormonal contraception or not, purely to see if this increased their risk of getting HIV.
Instead, the studies are likely to be observational or trials that have primarily been investigating other things.
This means there is potential that associations are being influenced by confounders. In short, other factors linked to contraceptive use, such as lifestyle behaviours, are themselves influencing the risk of HIV, rather than contraceptives directly.
What did the research involve?
The researchers built on the findings of a previous 2012 World Health Organization (WHO) review.
For the current review, they searched one literature database for English-language articles published from December 2011 onwards that included the terms "hormonal contraception", "HIV/acquisition", "injectables", "progestin", and "oral contraceptive pills".
They included studies that assessed hormonal contraceptives, included women without HIV at the study start, and were prospective in nature (following people over time).
Eligible studies were also required to have followed up at least 70% of their participants, have adjusted at least for a woman's age and condom use (to try to minimise confounding from these factors), and been conducted in a low- or middle-income country.
Separate researchers individually assessed the methods and quality of the eligible studies, and extracted data.
What were the basic results?
A total of 12 studies met the criteria for being included. All of these studies were conducted in low- or middle-income African countries.
These studies included large numbers of women, from between 400 to more than 8,000, and lasted between one and three years.
What were the studies investigating?
Three of the 12 studies included were observational studies designed specifically to examine any connection between contraception and HIV, while the other studies included women taking part in trials investigating interventions for HIV prevention.
Who was involved in the studies?
Most of the the 12 studies included looked at women aged 25 to 40 in the general population, while two looked specifically at women at high risk of HIV (commercial sex workers or women whose partner was HIV positive).
What contraceptives did the studies examine?
Some of the studies looked at women taking oral hormonal contraception (either combined pill or progestogen only).
In some the women were taking the injectable progestogen depot medroxyprogesterone acetate, and in the remaining studies the women were taking another type of injectable progestogen (norethisterone enanthate).
Most trials compared these hormonal types of contraception with a non-hormonal method of contraception, or with no method of contraception at all.
What were the specific results for the contraceptive injection?
Pooled results of 10 studies of depot medroxyprogesterone acetate found it was associated with a 40% increased risk of HIV (hazard ratio (HR) 1.40, 95% confidence interval (CI) 1.16 to 1.69).
This risk was slightly lower when restricted to only the studies of women in the general population (pooled HR 1.31, 95% CI 1.10 to 1.57) rather than those at high risk of contracting HIV.
There was no evidence of an increased risk of HIV in women taking the other injectable progestogen, norethisterone enanthate (pooled HR 1.10, 0.88 to 1.37); nor was there any increased of HIV risk found from the use of oral contraceptive pills (HR 1.00, 0.86 to 1.16).
How did the researchers interpret the results?
The researchers concluded their findings "show a moderate increased risk of HIV acquisition for all women using depot medroxyprogesterone acetate, with a smaller increase in risk for women in the general population.
"Whether the risks of HIV observed in our study would merit complete withdrawal of depot medroxyprogesterone acetate needs to be balanced against the known benefits of a highly effective contraceptive."
This is a well-conducted systematic review that tried to identify all studies investigating the possible link between hormonal contraceptive use and HIV.
It did not find an association between HIV risk and oral hormonal contraceptive use, nor with one type of injectable progestogen contraceptive.
But it did find an increased risk of HIV in studies where women used a commonly used injectable form of contraception called depot medroxyprogesterone acetate.
The review had strict inclusion criteria, but the possibility of selection bias and confounding from other factors still cannot be ruled out.
Only three out of the 12 studies directly set out to look at whether hormonal contraceptive use was linked to HIV. And these were still observational studies, meaning the women chose their method of contraception.
The other nine studies were not designed to look for this association.
As women in all of the 12 studies included chose their method of contraception, this could mean there are other differences – such as health and lifestyle – between the women who chose to use this type of contraception and those who chose to use non-hormonal methods. So the contraception may not have been the sole or direct cause of the link.
Two of the studies also included high-risk women, such as commercial sex workers or women whose partner was HIV positive. Exclusion of these studies decreased the association between depot contraceptive injection use and HIV, although the link remained statistically significant.
As the researchers themselves acknowledge, the studies can't say whether the association between hormonal contraception and HIV is "causal". And this is crucial to bear in mind when looking at this review.
Other limitations of this research
- As the authors also say, it is difficult for them to be sure of the timing of contraception use in relation to subsequent HIV infection.
- Although the studies included contraceptive methods that are used in the UK, none of these studies were UK-based and all were conducted in sub-Saharan Africa. The prevalence of HIV in these countries is much higher than it is the UK, so the baseline risk of contracting HIV is already much higher than it would be in the UK. The 40% risk increase with the depot injection is a relative increase of what would comparatively be a very small baseline risk in the UK.
Contraceptive injections such as Depo-Provera are extremely effective – estimated to have a failure rate of less than 1 in 330. But they provide no protection against sexually transmitted infections.
Only barrier methods such as condoms protect against HIV and other sexually transmitted infections, such as chlamydia and genital warts.
Talk to your GP if you're unsure you're using the most effective and convenient contraceptive for you.
Analysis by Bazian. Edited by NHS Choices. Follow Behind the Headlines on Twitter. Join the Healthy Evidence forum.