Reading the original paper, they compare Botswana and S Africa and find a correlation between seroprevalence and viral load. There is not good evidence to suggest this is causative however--this may be, for example, that as you get to higher frequencies of cases differently aged people are more likely to be infected and show different viral loads. A direct causal link between these two factors doesn't make a huge amount of sense, because (AFAIK) HIV is thought to sample all viral sequences within each host (so population-level selection should represent individual-level variation).
That's pretty typical for parasites (which include viruses).<p>Killing your host is never a good idea, the best parasites cause minimal disruption to their host.<p>All animals and humans are infected with countless parasites, but you never notice. For example <a href="https://en.wikipedia.org/wiki/Face_mite" rel="nofollow">https://en.wikipedia.org/wiki/Face_mite</a> (don't scratch).
<a href="http://www.davidbrin.com/givingplague.html" rel="nofollow">http://www.davidbrin.com/givingplague.html</a>
A very nice story of Brin based on the notion of a process of
infectious agents "watering down" over time.
"In theory, if we were to let HIV run its course then we would see a human population emerge that was more resistant to the virus than we collectively are today - HIV infection would eventually become almost harmless."<p>Because millions (or billions) of people would die, leaving only those with resistance.
> "Twenty years ago the time to Aids was 10 years, but in the last 10 years in Botswana that might have increased to 12.5 years, a sort of incremental change, but in the big picture that is a rapid change."<p>20 years ago, the ARV's and mixes weren't what they were 10 years ago and awareness + access to therapy also changed. Therapy initiation guidelines changed - which would directly impact time to AIDS.<p>Throw that statement away in relation to the meaning of the study results - except if this statement is there, it raises a concern about bias or mistakes in the study due to an insufficiently rigorous handling of the data and contexts/meaning.
Surely this process means that HIV is just as capable of evolving into something even more lethal over time, given the right environment? Given the relatively closed environment of a country like North Korea (for example) where diversity is restricted, its evolutionary path may take a different route than that observed by the University of Oxford research team?
As is so often the case, William Gibson came pretty close to predicting this:<p><a href="http://everything2.com/title/J.D.+Shapely" rel="nofollow">http://everything2.com/title/J.D.+Shapely</a>
What viruses "want" is to reproduce. Killing you is an unfortunate unplanned side effect.<p>If only we could negotiate and agree to allow them to reproduce and not kill us, we'd all get along much better.