A picture shows the hands of 104-year-old Chinese twins Cao Xiaoqiao and Cao Daqiao in 2009.A picture shows the hands of 104-year-old Chinese twins Cao Xiaoqiao and Cao Daqiao in 2009. (Reuters)

Predicting whether someone is genetically predisposed to live in good health to a very old age has moved a step forward.

In Thursday's online issue of the journal Science, researchers say their model of 150 genetic variants helped to predict exceptional longevity — living to late 90s or longer — among people in the study compared with people in the general population of North America and Europe.

The findings raise the possibility that people can learn in advance whether they have the potential to live to a very old age, though lifestyle choices and environmental factors remain important.

"Centenarians are indeed a model of aging well," a senior author of the study, Dr. Tom Perls of the Boston University School of Medicine and Boston Medical Center, told reporters.

"They very much compress their diseases or their disability towards the very end of their lives."

"Inheritability of longevity has been looked at, so genes do play a role," said Dr. Kenneth S. Kendler of the Department of Human and Molecular Genetics at Virginia Commonwealth University.

But so do other factors "such as driving motorcycles fast and smoking," said Kendler, who was not part of the research team.

The researchers identified a group of 150 genetic variants that they say can predict exceptional longevity with 77 per cent accuracy. The 77 per cent accuracy rate reported in this study is better than other groups have been able to do, Kendler noted.

The team found 19 genetic "signatures" in the subjects' DNA that were associated with diseases such as dementia and hypertension.

Scientists have long thought that the ability to live to 100 involves a relative lack of genetic variants associated with disease.

The new findings, however, suggest that is not the case. Instead, longevity-associated variants could protect or even cancel out some of those disease effects, said study co-author Paola Sebastiani, a professor of biostatistics at the Boston University School of Public Health.

Dent in Alzheimer's

Down the road, the signatures could tell a physician sooner whether a patient is at increased risk for a particular disease.

"I look at the complexity of this puzzle and feel very strongly that this will not lead to treatments that will get a lot of people to become centenarians, but rather to make a dent in the onset of age-related diseases like Alzheimer's," Perls said.

The researchers caution that genetic tests currently sold at U.S. pharmacies don't take into account other genes that could be cancelling out or adding to the disease's effects. They argue the implications of their model need to be better understood, and as such, the technology should not yet be relied upon.

To make their findings, the researchers compared disease-associated variants in 801 unrelated subjects enrolled in the New England Centenarian Study with 926 controls with the same Caucasian genetic backgrounds.

The study focused on Caucasians, tracking subjects going back as far as 1995. But the researchers say they plan to extend the study to other ethnic groups. For example, they would like to study people from Japan, a country that has a large number of elderly people.

The study was funded by grants from the U.S. National Institute of Aging and the National Heart Lung and Blood Institute of the National Institutes of Health.

The researchers have no financial interest in the research and are not planning to patent the technology. Rather, if people are tested for 150 genetic signatures, the New England team is planning a website to compute risk profiles and interpret the results.

With files from The Associated Press