In mid-November New York magazine ran a cover story on a group of 540 Ashkenazi Jews who, as the headline trumpets, carry "DNA You Wouldn't Believe." This group of happy oldsters, all of them over 95, might hold the key to extreme longevity, the article suggests. There is the 105-year-old stockbroker who still goes to work everyday and a peripatetic 109-year-old woman.

The article—"What Do a Bunch of Old Jews Know about Living Forever?"—highlights research from the Albert Einstein College of Medicine in the Bronx that points to several genes that might help explain the secret of a Methuselah-like existence. One, in particular, stands out, the Einstein researchers assert. The oldest of the old who bear a certain version of the cholesteryl ester transfer protein gene, (CETP I405V) end up with lower levels of the CETP protein but higher levels of the "good" cholesterol, and larger particles of these high-density lipoproteins (HDLs). The carriers of two copies of this gene variant—one from ma, one from pa— appear to have a lower risk of dementia, failing memories and heart disease.

There is a problem with this halcyon picture. A few weeks after the New York article hit the presses, a report by researchers from Rush University Medical Center in Chicago in a respected journal, Aging Cell, identified the CETP variant as a gene that may increase the risk of Alzheimer's. The same version of the same gene: in one study it slows dementia, in the other it promotes cognitive decline.

What is going on here? Two groups of well-regarded researchers come to the exact opposite conclusion. Welcome to the wild-and-wooly world of gene-association studies that attempt to tie DNA to disease. Unless the link between a gene and a disease or phenotype is a very strong one, researchers who try to follow up on initial studies often get different results the second or third time around.

In this instance, the researchers at Rush had decided to investigate this gene after reading a January study published in JAMA The Journal of the American Medical Association by the Einstein investigators. The JAMA article found that the CETP variant protected a diverse group of older people, aged 70 and older, not just the Ashkenazim, against failing memory and dementia. The JAMA study, though, extended earlier work by the Einstein researchers that showed better cognition and general health in the Ashkenazi super old that carry the gene.

The Rush team, which oversees two well-known longitudinal studies on aging and dementia, decided to see whether they could replicate the JAMA findings. Their analysis of 1,384  participants from the two studies found that the CETP variant was associated with a faster rate of cognitive decline and an increased risk of Alzheimer's. The Chicago researchers then did something that the Einstein team hadn't. They went to pathology information collected on the brains of 590 of the deceased study participants. They found there that those with the relevant CETP variant had a greater density of the plaques characteristic of Alzheimer's than others who didn't.

So who is right? Maybe no one at all. A leader of the Rush team points out the possibility that both results may be due to chance "They got an effect, they reported it; we got an opposite effect, we reported it," said David Bennett, a professor of neurological sciences at Rush and one of the co-authors on the paper. "Maybe the truth is in the middle, which is no effect."

Another possibility may relate to differences in who the two studies surveyed—Rush looked at a population of European ancestry; the 523 Bronx subjects in the JAMA study came from a mixed racial and ethnic background. The gene variant might have different effects in different groups. Could that be why the Ashkenazi centenarians shine?

Maybe. But it still might be a long shot. Findings that the CETP variant enhances cognition and general well being in old people have been equivocal. Some studies have shown benefit, others haven't. Desultory outcomes in gene association studies often mean that a single gene may have a relatively weak effect, if any. "It's important that we try hard to understand how genes affect diseases of later life but we need to moderate our expectations of what we can expect to find," says Tom Kirkwood, a biologist from Newcastle University in England. (See “Why Can’t We Live Forever” by Kirkwood in the Sept. 10 Scientific American.)"All the indications are that we're most unlikely to find that single genes make big contributions, and we're likely to need to develop rather more sophisticated ways of probing and interpreting the data than we have at present."

The CETP dispute is not the only one of recent vintage. Earlier this year researchers from Boston University retracted a paper published in Science that purported to have found 150 genetic variants linked to extremely long lives, though the investigators said they plan to submit their results to another journal.

The race to probe the genomes of the very old continues. The Archon Genomics X PRIZE, which kicked off this year, will award $10 million to the first team to sequence the genomes of 100 centenarians in an effort to unlock clues to healthy aging.

Interest in CETP for improved cognition in the elderly has not flagged either. Several drugs under development would inhibit production of the CETP protein in the same way as the gene variant purportedly tied to longevity does. The rationale: Lowering CETP results in higher HDL, which may protect against heart disease. "The question is if it will be relevant to cognition or just to cardiovascular endpoint," wonders Nir Barzilai, one of the Einstein researchers.

Drug companies like Merck and others have sunk millions into development of CETP inhibitors. Could cognitive decline be a side effect of these drugs? Probably not, but it still might be a question worth asking.

Recruitment for a drug trial to test effects of these drugs on cognition would be fascinating: Subjects would be recruited to see if a CETP inhibitor improves mental functioning or whether it may speed decline into dementia. Any volunteers?