• Question: During your research, how do you separate the many factors that are believed to lead to cancer, such as, smoking and excersise?

    Asked by 05braceye to Wei on 14 Jun 2011.
    • Photo: Wei Xun

      Wei Xun answered on 14 Jun 2011:

      Hello!

      Here we get quite technical. Of course you’re totally right, many things that we do that might affect our health are related. Suppose someone who smokes heavily is also unlikely to be an exercise enthusaist. So how do we know which is which? This is what we call an “interaction”.

      So, firstly, it would help if you have a clear picture in your mind. Ie if you already have an idea of whether it’s mainly the smoking that’s causing say lung cancer, then you go and test it in this way. You can look at people who exercise the same, but smoke in different amounts so you can see the link.

      Secondly, we can use statistics. There are many clever statstician who have developed different mathematical models to be used just for this situation. It crosses over into philosophy as well, just think, what exactly causes something to happen?

      But when the two things are so almost completely wrapped up with each other then there’s no way for us to guess what is really happening, so we have to wait for an opportunity for them to be seperated, eg by looking in other populations, ie where people don’t smoke.

Comments

  • Photo: 05braceye

    05braceye commented on :

    Thanks, so because these carcinogens are so closesly related, has anything been definately proved to lead to or cause cancer? Because if several factors are involved, surely it is unknown whether or not something is directly involved?

    • Photo: Wei

      Wei commented on :

      So this is where research from other fields will come to help us. In the case of tobacco smoke, animal and cellular experiments have shown that the compounds in it actually give rise to lung tumours, and there are believable pathological mechanisms for the processes.

      In epidemiology, there are a set of things you need to consider before we can infer that one thing causes a disease or not: http://en.wikipedia.org/wiki/Epidemiology#Bradford-Hill_criteria.

      However, this is Masters degree territory, if you’re interested I would recommend that you read a good epidemiological textbook such as this one: Modern epidemiology / Kenneth J. Rothman 2008. ISBN: 9780781755641 It’s available in the British library if you can make it there.