AN EPIDEMIOLOGIC finding that many said did not make sense is proving to be more than a statistical fluke.
Skepticism greeted reports last year of a strong association between Alzheimer's disease and the genes for apolipoprotein E (apo E), which were thought to be important only in cholesterol metabolism (JAMA. 1994;271:89-91). Critics scoffed that apo E could not play a role in neuronal death because it was never seen in neurons.
But during this past year, dozens of studies around the world have confirmed that the apo E4 allele confers the greatest risk of developing late-onset Alzheimer's, the relatively rare E2 allele seems to be protective in most but not all ethnic groups, and risk with the most common E3 allele lies in the middle.
One new study suggests that the E4 allele is the strongest predictor of which patients with mild memory problems will go on to develop Alzheimer's