A multifactorial model was used to examine whether risk group identification was independently significant in the presence of confounding variables such as age and sex.