If potential risk factors are then related to estimates of later multiple disease states, it can easily induce misplaced confidence in the validity of the conclusions about which risk factors were responsible for what disease burden. I am especially concerned about reliance on manipulating disparate surveys that rely on people remembering and reporting what they consume, and which may not always adequately adjust for key confounding factors. Its startling findings on for example the apparent high risk of death when not enough fruits or nuts are consumed, may be an artifact of the methodologies chosen by the statisticians who are the lead authors of the study. Last month I expressed doubts about the validity of the gigantic Global Burden of Disease study whose initial results were published in December in The Lancet (1,2). TheseĬontributions now continue here as a regular column. Philip James has written a series of ten contributions identified as 'As I see it', published monthly on the Association's home page since March 2012.
0 Comments
Leave a Reply. |