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 sugge

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).