HIV Risk and Injectable Contraceptives: What Do We Really Know?
By Fiorenzo Conte
The short answer would be: there is some (weak) evidence that injectable contraceptives might increase the risk of HIV infections, however, none of the findings is conclusive.
A large study conducted in several African countries reveals that hormonal contraceptives (more specifically injectable depot-medroxyprogesterone acetate DMPA) might double the risk of HIV contraction and infection. The study, whose target population was constituted by 3800 discordant couples (couple in which only one partner is infected with HIV1), points to increased susceptibility of uninfected women using DMPA to HIV infections. Similarly, according to the research, women who make use of the oral contraceptives are twice more likely to infect their respective partners vis-à-vis women who do not use injectable contraceptives . As injectable contraceptives constitute one of the most effective measure in reducing undesired pregnancy and therefore maternal mortality, these findings would give reasons for concerns. However, before jumping to definitive conclusions, several methodological concerns of the study should be mentioned . When taken into consideration, these methodological issues stress that the findings are far from conclusive and further research is needed in order to make any conclusive statement about the relationship between hormonal contraceptives and HIV infection.
The first concern has to do with internal validity. The study was not designed specifically to test the impact of the use of hormonal contraceptives on HIV infections therefore presents a series of pitfalls. One of them is that use of contraceptive was self reported and not backed up by clinical check. Further, the sample size of those women using DMPA and HIV positive was very very small: only 10 women using DMPA were in fact infected. Similarly, only 11% of the women in the sample used DMPA. The number of observation is really too small to have a high degree of confidence about the results. Lastly, the risk of selection bias is concrete as the treatment (in this case the use of injectable contraceptive) has not been randomly assigned. It follows that women who decided to use the DMPA could have been systematically different from those who do not use them and therefore any comparisons between the two groups would not stand.
The second concern has to do with external validity. Discordant couples constitute a small subgroups of the population in most of the countries in the studies (think the only 10% of the couples in Kenya are HIV positive and of these, 6 out 10 are discordant couples according to the 2007 Kenya AIDS Indicators Survey) and therefore it might be the case that even if the findings apply for this subgroups it might not be valid for the general population.
On the basis of the discussion above, the findings are nothing else but chance observations and should be treated as such. Unfortunately, the tendency of the media is to overreact to such inconclusive studies and this might be detrimental when it comes to policy decisions or choice made by individuals. However, before any further evidence is presented the benefits of the DMPA far outweigh the risks associated with it.