A new study from a multi-center group of researchers led by Drs. Howard Federoff of the Georgetown University Medical Center and Mark Mapstone of the University of Rochester, has given some hope for utilizing a blood-test panel of 10 phospholipids to predict the development of Alzheimer s disease (AD) or its precursor, mild cognitive impairment (MCI). Unfortunately, the authors somewhat optimistic appraisal of their own work is a bit overly enthusiastic and not completely justified by cold statistical analysis.
The researchers, seeking some hint of a rewarding area of exploration for the biochemical origin[s] of AD, decided upon phospholipids as an exemplar of cell membrane breakdown products. They analyzed numerous variants of these substances in the blood from different groups of study subjects, and came upon the particular 10-member panel of phospholipids as showing some prognostic promise. They then studied a group of 575 people over age 70, and over the course of five years, AD or MCI was detected in 28 of those. They tested the 10-lipid panel and found a high correlation between low levels of that group and the AD/MCI diagnoses, when compared to the normal control group. Working backwards with this knowledge in hand, studying the other non-AD/MCI patients in the original group, they determined that finding a significantly low level of the 10-lipid panel predicted AD/MCI with a likelihood of 90 percent.
"We were surprised," Dr. Mark Mapstone told CNN Health s Elizabeth Cohen. "But it turns out that it appears we were looking in the right place."
Not so fast, maybe: According to MedPage today s analysis, [T]he data did not appear to fully support that optimism. If the study cohort's 5% rate of conversion from normal cognition to mild impairment or Alzheimer's disease is representative of a real-world screening population, then the test would have a positive predictive value of just 35%. That is, nearly two-thirds of positive screening results would be false.
ACSH s Dr. Gil Ross had this comment: The prospect of many millions of AD victims over the coming decades is a huge public health problem as our population ages. We have nothing to offer these people, and worse, we have no way to prevent it or even delay its onset. We need some way and it s got to be rapid, cheap, and not too invasive to predict, accurately, the likelihood that an individual will develop AD so that pharmaceutical researchers can do clinical studies on various therapies to see which ones might work, since waiting for clinical symptoms is definitely too late. This 10-lipid panel is far from the holy grail, as the authors seemed to perceive it, but it may well help to make future trials more focused by getting a higher percentage of future AD patients in the mix instead of just using a random sample of the general population.