Posted 3/13/2010 3:22 pm
"In November 2007, the World Cancer Research Fund and the American Institute for Cancer Research published their second expert report, entitled Food, Nutrition, Physical Activity and the Prevention of Cancer: a Global Perspective.[8] This report took five years to prepare and reviewed more than 7,000 studies published worldwide.[9] Among the recommendations of the report is that, except for very rare occasions, people should avoid eating ham or other processed meats -- cured, smoked, salted or chemically preserved meat products such as bacon, hot dogs, sausage,[10] salami,[11] and pastrami. The report states that once an individual reaches the 18-ounce (510 g) weekly limit for red meat, every 1.7 ounces (48 g) of processed meat consumed a day increases cancer risk by 21%.[9]"
Yeah, "if" being the operative term - these sort of meta-studies have a well established reputation for turning up remarkable results that nobody can ever replicate later.
Posted 3/13/2010 4:20 pm
Actully I WISH that study was true but if it was then 90% of the country would be dead!!! Howevever there is some truth to the study. There is NO question meat increases cancer risk but how much?
But that's the point - this *isn't* a study - it's a meta analysis - and, as I mentoioned a couple of posts above, they have a terrible history of producing irrepeatable results. The basic problem is that you need to normalize all the data - and this is an extremely hard thing to do without the risk of introducing biases.
This is, obviously, not to say that all results from this type of work are automatically invalid - just that they tend to have a much lower level of confidence associated with them than studies that are carried out with a controlled sample set. When you are dealing with cancer data, this is an ever bigger problem, since all the data is so noisy in the first place.
Which part of "meta analyses have a terrible history of producing " didn't you understand?