HIV has become more virulent over time, not less, European study finds

Local conditions may drive viral evolution in opposite directions
This article is more than 9 years old. Click here for more recent articles on this topic

The largest cohort study ever to look at CD4 count and viral loads in HIV-positive people around the time of diagnosis has found evidence that HIV, at least in Europe, has become more virulent over time. The average time taken to reach a CD4 count below 350 cells/mm3 has halved over the last 25 years, researchers calculate.

The study of 15,875 patients within the pan-European CASCADE cohort finds that the estimated CD4 count after seroconversion (a measure of how strongly the acute phase of HIV infection attacks the immune system) fell from 770 cells/mm3 in 1979 to 570 cells/mm3 in 2002.

The average ‘set point’ viral load – the rate at which HIV reproduces in the absence of treatment, after the initial burst of viral replication – increased from 11,200 copies/ml in 1980 to 31,000 copies/ml in 2002 (with a possible slight decline to 25,500 copies/ml by 2008).



The power of bacteria or viruses to cause a disease. Different strains of the same micro-organism can vary in virulence.



The transition period from infection with HIV to the detectable presence of HIV antibodies in the blood. When seroconversion occurs (usually within a few weeks of infection), the result of an HIV antibody test changes from HIV negative to HIV positive. Seroconversion may be accompanied with flu-like symptoms.



In discussions of consent for medical treatment, the ability of a person to make a decision for themselves and understand its implications. Young children, people who are unconscious and some people with mental health problems may lack capacity. In the context of health services, the staff and resources that are available for patient care.


The process of viral multiplication or reproduction. Viruses cannot replicate without the machinery and metabolism of cells (human cells, in the case of HIV), which is why viruses infect cells.

set point

The viral load that the body settles at within a few weeks to months after infection with HIV. Immediately after infection, a person’s viral load is typically very high. After a few weeks to months, this rapid replication of HIV declines and the person's viral load drops to its set point. A higher viral set point suggests that, in the absence of treatment, disease will progress faster than in a person with a lower set point. 

The rate at which CD4 counts declined increased during the same time period. This meant that the average time taken for the CD4 count of a person who seroconverted in 1980 to fall below the current UK treatment-initiation threshold of 350 cells/mm3 was seven years; by 2002 this had halved to 3.4 years.

Contrast with African study

These figures differ strikingly from a recently published study (Payne et al.) which finds that HIV’s replicative capacity has declined over time, at least in southern Africa. This study received wide publicity in the world’s media after being reported by the BBC on World AIDS Day.

In this study, which was originally reported by aidsmap from the HIV Vaccine Conference in 2013, researchers from the University of Oxford found that HIV from Gaborone in Botswana, where HIV prevalence peaked in the year 2000, reproduced 11% more slowly than virus in Durban, where HIV prevalence peaked in 2010.

In addition, the average viral load in the untreated population was 15,000 copies/ml in Gaborone and 29,000 copies/ml in Durban – though it appeared to have started falling very recently there too.

Lead researcher Philip Goulder also found that HIV in Gaborone had more resistance mutations against human immune system proteins. The research group therefore speculated that HIV, in trying to evade human immune defences, was mutating over time into a less virulent form.

More on the European study

The CASCADE study finds the opposite. Using the three measurements described above, it found evidence of a considerable increase in HIV’s virulence and replicative capacity between 1979 and 2002. After this, the rate of CD4 decline and set-point viral load reach a plateau and there is no significant change up to 2008, the last year in which figures for this study were measured. There is evidence of a slight decline in set-point viral load after 2002, as detailed above.

The researchers performed sensitivity analyses to rule out other influences on CD4 count and viral load. In particular, they:

  • Measured the figures only in white gay men (because they were more likely to be European and have had a similar treatment history and viral subtype).

  • Measured the figures only in patients who were quite definitely diagnosed within three months of infection (because this would mean the CD4 count at seroconversion and the setpoint viral load figures were more accurate).

  • Measured the figures only up to three years after infection (to reduce any bias caused by the people with the fastest CD4 count declines going on treatment first).

These sensitivity analyses made no difference to the measurements of set-point viral load and CD4 count at seroconversion. However, whereas the rate of CD4 count decline appeared to plateau from 2004 onwards in the whole cohort, it continued to worsen in white gay men and even more notably in people diagnosed within three months of infection.

If these virulence figures are true, the findings have considerable implications for testing and treatment policies – as they mean people have a shorter time window in which to get tested and go on treatment before they get sick. This “substantially shortened time…emphasises that individuals cannot afford to delay testing for HIV,” comment the researchers.

In addition, the study has prevention implications, because the researchers calculate that the higher viral load set-point implies a 44% increase in the average infectiousness of untreated patients in CASCADE.

Why do these two studies find such different results?

One possibility is that we are just starting to see a decline in virulence in CASCADE, and there is some indication of that, not only in set-point viral load, but also in trends that are as yet not statistically significant, in the direction of a higher CD4 count at seroconversion.

However, the European epidemic in gay men, at least, is no younger than the epidemic in Botswana, so there is no apparent reason it should be happening just now.

Richard Jefferys, of the New York-based Treatment Action Group, argues in a blog that the Gaborone/Durban study results may not be generalisable. The replicative capacity of HIV in the samples was measured at only one time point, and there is no evidence that this has any effect on subsequent CD4 declines in chronic infection, he said.

In addition, the African study is very small: HIV replication capacity was only compared in samples from 63 people in Gaborone and 16 in Durban; this contrasts with data from tens of thousands in CASCADE.

However, the most likely reason is that the differing characteristics of the African and European epidemics are causing HIV’s virulence to move in opposite directions. HIV is an immensely genetically variable organism, and one that adapts very fast to local conditions. To conclude that a decrease in virulence in one area applies globally is like concluding, by analogy, that because rhinos are endangered in southern Africa, foxes must be in Europe. A 2007 modelling study found that HIV’s viral load will tend to stabilise at the point at which the virus is most readily passed on locally, for evolutionary reasons.

In a heterosexually spread epidemic where, until recently, few people have been on treatment, it makes evolutionary sense for HIV to become somewhat less virulent. This is because its ‘victims’ need to stay alive long enough for the virus to get itself transferred into another body. Of course, the viral selection strategy is a result of natural selection, not intelligent adaptation. Viral variants that replicate at fairly low levels in their hosts will eventually be transmitted, as low viral load means that the host lives longer.

In an epidemic where HIV spreads more rapidly, for instance between gay men, but where a majority of people are on treatment, evolutionary pressure will tend to push the virus towards greater virulence: HIV must be transmitted before people start treatment, favouring transmission from people with higher viral load. These viral variants cause more rapid disease progression. There is no evolutionary penalty for faster progression to AIDS for these viral variants because in this highly-treated population, fewer people die before the virus is passed on.

It is notable that the viral set point seen in 2002 in the CASCADE study (31,000 copies/ml) is near the optimum viral load calculated for the most efficient transmission in the 2007 model (33,000 copies/ml).

In other words, both studies could be right: HIV could currently be weakening in one part of the world, while becoming stronger in another.


Pantazis N et al. Temporal trends in prognostic markers of HIV-1 virulence and transmissibility: an observational cohort study. Lancet HIV1:e119-126. See for abstract. 2014.

Payne R et al. Impact of HLA-driven HIV adaptation on virulence in populations of high HIV seroprevalence. Proc Nat Acad Sci, early online edition. See for full text. 2014.