Author: Admin (---.cu27.vnnyca.adelphia.net)
Date: 06-02-02 16:44
Debbie,
This myth about whether smokers and non-smokers are more likely to "get" sarcoidosis is an excellent example of exactly what is wrong with a lot of medical research these days.
Too much medical research is based on what is called "epidemiological" studies. Instead of trying to get an understanding of what endocrinological factors are behind a disease, the epidemiologist gathers together a lot of statistical data from a census, or from a hospital's records, and tries to make inferences based (usually) purely on the statistical (epidemiological) data.
I, too, was trained in statistics. But my professor introduced the course with a book called "How to Lie with Statistics". Effectively it showed the many ways in which statistics can be misused in order to prove one's own point. We were taught to make sure that we understood the problem before beginning to draw inferences from the data.
Now I am not accusing the medical profession of trying to prove something that is false. They are not. They are well intentioned. It's just that they try to draw statistical inference without having all the possible factors confounded (a mathematical term for 'removed') from their data.
For example, for decades we have had paper after paper published saying that steroids are good for sarcoid patients, that steroids improve things. But when those studies are evaluated by expert epidemilogists, such as the Oxford study, or this more recent review for JAMA, they are found to have forgotten about one or other important factor that might have totally changed their conclusions.
In fact, of the 150 epidemiological studies of prednisone treatment that were subjected to review, only 8 provided usable data. How many patients suffered following the recommendations of the other 142 ?
Another study purported to show that there was an "Increased risk for cancer following sarcoidosis". I looked at this study myself, and realized that the researchers had made no attempt to distinguish patients who might have been treated with Methotrexate or Steroids (or indeed any drugs likely to act as carcinogens) from those people who were not sick and who had not been exposed to medication. I wrote to the main author, Dr Askling, questioning why they had not thought this to be significant. His response was that they "didn't have that data available to them". So why bother to do the study at all? Or why not confess this problem with the study results? But there were no disclaimers, just a headline which would unecessarily scare many sarcoid patients and their doctors 
So I finally get to your question about smokers and sarcoid. Sorry it took so long, but I had to get that off my chest 
Well, there are two big fundamental problems with the smoking studies that I have seen. They haven't considered sarc as a lifelong pre-disposition (many patients can't be said to 'get' sarc) and they have not considered the possibilty that when sarc patients were offered cigarettes they found them to be especially systemically irritating, and therefore never took up the habit.
Other patients seem to develop the sarc genes later in life, in much the same way that gene mutations that allow cancer to develop. Maybe these patients are more frequently smokers, and represent a different subpopulation? Until Doctors recognise these confounding factors we will never know if they do. Nothing that I have read goes anywhere near 'proving' that non-smokers are more likely to 'get' sarcoidosis. This is a myth resulting from bad science.
Furthermore, recent laboratory studies have shown that smoking increases the amount of ACE generated in the granuloma of sarcoidosis patients, indicating that smoking seems to increase the inflammation in the lungs of sarc patients...
..Trevor..
|
|
