Thursday, May 26, 2005

Skeptics join with crackpots?

I haven't posted recently due to spending a wonderful vacation in Umbria (the region of Italy roughly in the middle between Florence and Rome) without internet connectivity and then coming back to find IUB in exam week mode and trying to catch up with physics (see also my post at the coffee table).


This week, we had Serge Lang as a visitor. You'll probably know some of his roughly fourty math books (many excellent text books). In three days, he gave four talks, two on the abc-conjecture (a conjecture in number theory that can be described elementarily and implies Fermat. The corresponding theorem (with simple and elementary proof) about polynomials has been only found only in the past 30 years. Then there was a colloqium on ??? (people later told me it prominently featured the heat-kernel) which I had to miss due to an exam. And there was yesterday...


I already got a sneak preview of yesterday's talk over lunch on Tuesday so I could prepare myself with some web search for the Q&A session. The other thing Lang is known for is that he gets into heatet fights with people. This time, he claims (and he is not alone with this claim) that AIDS (or what is called AIDS) is not caused by the HI-virus. HIV is a harmless virus that does not cause the syndrome called AIDS.


His version roughly goes like this: What is called AIDS are really different illnesses in the different risk groups: In the homosexual community it is caused by the toxic effects of some drugs called "poppers" (amyl nitrite) that became popular in that community in the early 80ies and all studies with the typical San Francisco crowd of participants measured only their preference for those drugs. Poppers are sniffed but also intravenous drug users are affected by AIDS because of the toxic effects of these drugs. And what is observed in Africa is really only malnutrition. And of course in all groups the prescription drugs like ATZ with their many side effects cause many of the symptoms of AIDS.


Matters are complicated by the "official" definition of AIDS in the US: There the working definition is that you have one of a list of roughly 30 auxiliary illnesses and are HIV positive. So by definition, you cannot have AIDS without being HIV positive. However, he ignores that it is an empirical observation that these illnesses and especially some count of T-helper cells is a strong predictor of HIV positivity.


Furtheremore, the test (actually there are several different ones) for HIV is not perfect and gives false positives. And here the laws of statistics apply. Just to make up some numbers, assume that if you do not carry the virus, there is a 5% chance that the test tests you positive. But only 0.3% of people carry the virus. So if I take the average person on the street and test her and the test is positive, chances are still 95% that she does not have the virus. It's just that it's not 99.7% anymore and a second independant test will give you a much better anser. So Lang can quote important people with statements that if the test is positive it does by far not mean you have the virus.
Finally, it is notoriously difficult (and so far has been impossinble) to isolate the virus so you cannot just look at it under the electron microscope.


As you can imagine there was a heated discussion and Lang wasn't always the person who tried to keep the argument rational. There was some other guy in the audience who identified imself as a medical doctor and strongly refuted Lang's claims. However after the talk, it turned out that this guy is from another camp AIDS/HIV critics that denies even the existence of the HI virus and claims that HIV/AIDS is not infectious. You think this claim can easily be refuted? Well, it turns out that such studies are very hard to do because you cannot just go around and try to infect people with HIV (or do even simpler things like treat HIV patients with a placebo instead of ATZ) for ethical reasons. Animal studies are also not 100% clear cut because human HIV does not infect animals and has to be modified to infect apes. So the upshot is it is not easy to argue with these people especially if you are not an expert.


Lang of course is a mathematician, but there are many other people around and not all of them have a sound mathematical background and can be confused by citing some numbers although everything looks ok if you really know how to interpret conditional probablilities. For example Lang cited newspaper articles of a woman that has been killed by giving hear anti-AIDS drugs. OK, this is sad but this might happen. What he failed to mention was of course what would have happend to that woman (on the average of course) if she would not have been treated with these drugs. If I collapse here in my room I want the ambulance drive take me to the hospital quickly even if this involves an increased risk for me to die in a traffic accident.


I will not give you any references to these anti HIV/AIDS critics. You can google for them yourself they are not hard to find. However, I would like to mention a few pages where the establishment (or orthodody as Land calls them) answers to these claims:



  • A collection by the Robert Koch Institut (German Federal Health Institute)

  • A short text.

  • A longter one form Skeptic Magazine

  • And finally the Wikipedia entry on the controversy (marked controversal):
  •