Group of genes may predict longevity with 77% accuracy
July 2, 2010
Courtesy of Boston University Medical Center
and World Science staff
Re*search*ers have iden*ti*fied a group of genes that they say can be used to pre*dict with 77 per*cent ac*cu*ra*cy whe*ther people will live ex*cep*tion*ally long.
The find*ing is a break*through in un*der*stand*ing the role of genes in de*ter*min*ing hu*man life*span, ac*cord*ing to the sci*en*tists.
They stressed that en*vi*ron*ment and family his*to*ry are al*so fac*tors in healthy ag*ing. Yet the study “shows that ge*net*ic da*ta can in*deed pre*dict ex*cep*tion*al longe*vity with*out knowl*edge of any oth*er risk fac*tor,” wrote the researchers in their re*port, pub*lished July 1 on*line by the re*search jour*nal Sci*ence.
“Fur*ther in*ves*ti*ga*t*ion is needed to un*der*stand how and why these vari*ants col*lec*tively pre*dis*pose for ex*cep*tion*al longe*vity,” added the in*ves*ti*gators, from Bos*ton Uni*vers*ity Schools of Pub*lic Health and Med*i*cine and Bos*ton Med*i*cal Cen*ter.
The team con*ducted a genome-wide gene as*socia*t*ion study in cen*te*nar*ians, who of*ten don’t suf*fer age-related dis*abil*i*ties un*til well in*to their mid-nineties. Re*search*ers led by Bos*ton Uni*vers*ity’s Paola Se*bas*tiani and Thom*as Perls built a ge*net*ic mod*el that in*cludes 150 ge*net*ic vari*ants, and which they said could be used to pre*dict wheth*er a per*son lived to the late 90s or old*er.
The anal*y*sis al*so iden*ti*fied 19 ge*net*ic clus*ters or “ge*net*ic sig*na*tures” of ex*cep*tion*al longe*vity that they said char*ac*ter*ized nine in ten cen*te*nar*ians stud*ied. The sig*na*tures cor*re*lat*ed with dif*fer*ences in the prev*a*lence and age-of-onset of dis*eases such as de*men*tia and hy*per*ten*sion, and may help iden*ti*fy key sub*groups show*ing healthy ag*ing, the au*thors said.
The team found that 45 per*cent of the old*est cen*te*nar*ians – those 110 years and old*er – had a ge*net*ic sig*na*ture with the high*est pro*por*tion of longe*vity-as*sociated ge*net*ic vari*ants. “These ge*net*ic sig*na*tures are a new ad*vance to*wards per*sonalized ge*nomics and pre*dictive med*i*cine, where this an*a*lyt*ic meth*od may prove to be gen*er*ally use*ful in pre*ven*tion and screen*ing of nu*mer*ous dis*eases, as well as the tai*lored uses of med*ica*t*ions,” said Perls.
The re*search*ers de*vel*oped a new sta*tis*ti*cal ap*proach to an*a*lyze ge*net*ic da*ta from more than 1,000 cen*te*nar*ians and sev*er*al con*trol groups, and to iden*ti*fy vari*ants most pre*dictive of be*ing cen*te*nar*ians or not. “The meth*odology that we de*vel*oped can be ap*plied to oth*er com*plex ge*net*ic traits, in*clud*ing Alzheimer’s dis*ease, Parkin*son’s, car*di*o*vas*cu*lar dis*ease and di*a*be*tes,” Se*bas*tiani said.